SETIUD‘IEs'VA’; .SPEcmoscom , 0F IONIC VSOLVATl-ONT ‘APHICI GAS CHROMATOGR STUDIES OF CONVULSANT TE PART] PART u: ,TRAZOLES'; ‘ E ._ mm Asmr mm m __ m. , . . M . Date 0-7 639 This is to certify that the thesis entitled PART I: SPECTROSCOPIC STUDIES OF IONIC SOLVATION PART II: GAS CHROMATOGRAPHIC STUDIES OF CONVULSANT TETRAZOLES presented by Robert G. Baum has been accepted towards fulfillment of the requirements for Ph.D. deg?ein Chemistry fire a?” Major professor March 12, 1976 ABSTRACT PART I: SPECTROSCOPIC STUDIES OF IONIC SOLVATION PART II: GAS CHROMATOGRAPHIC STUDIES OF CONVULSANT TETRAZOLES BY Robert G. Baum The solvation of the lithium ion by acetone was studied in acetone-nitromethane solutions by far—infrared, Raman, 7 3S and Li and Cl nuclear magnetic resonance spectroscopic techniques. It was confirmed that the 390-cm-l far-infrared acetone band is split by the lithium ion and that a 369-cm-l far-infrared band, which was attributed by other investi- gators to a Li+-nitromethane vibration, is due to the vibra- tion of acetone molecules in the inner solvation shell of the lithium ion. It was determined that the frequency of the lithium ion vibration, in nitromethane solutions, is strongly dependent on the nature of the counter ion. Studies of the Li+-acetone—nitromethane system by several different experimental techniques indicate that the primary solvation shell of the Li+ ion consists of four acetone molecules. From Raman spectral data of the above system at varying compositions, approximate values of the equilibrium constants for the stepwise solvation reaction were calculated. The values obtained were Kl = 19.1, K2 = 2.5, K3 = 1.3, and K4 = 0.6. Lithium-7 nuclear magnetic resonance measurements showed that acetone solvates Robert G. Baum the lithium ion much more strongly than does nitromethane. Results from chlorine—35 nuclear magnetic resonance studies indicated that in solutions containing less than four acetone molecules per Li+ ion, the vacant position in the inner solvation shell is occupied by C10; ion in preference to nitromethane. The influence of a weak complexing agent, pentamethyl— enetetrazole, on the Li+ClOZ ion pair formation was in— vestigated. It was determined that the degree of inter— action with the lithium ion was acetone > pentamethylene- tetrazole > C102. Therefore, pentamethylenetetrazole dis— places ClO_, but not acetone, from the inner solvation shell of the lithium ion. Gas chromatography was used as a technique for the analysis of the cyclopolymethylenetetrazoles. The relative retention times were very nearly the same for trimethylene— tetrazole, tetramethylenetetrazole, and pentamethylene— tetrazole, but varied considerably as the number of methylene groups is increased. Working curves were obtained for the cyclopolymethylenetetrazoles and were found to be linear over the concentration range from millimolar to 0.1 g. At higher concentrations the curves deviated from linearity. These compounds were analyzed routinely at the 50—100 ppm level, and the technique is capable of determining concen— trations of less than 10 ppm. The gas chromatographic technique was also employed to study the distribution of these tetrazoles between Robert G. Baum aqueous solutions and carbon tetrachloride. For penta— methylenetetrazole, hexamethylenetetrazole, heptamethylene— tetrazole, and 8-tgrt-butylpentamethylenetetrazole, the distribution ratio was found to be concentration dependent. Cryoscopic measurements were carried out on aqueous solu— tions of the cyclopolymethylenetetrazoles. The results in- dicated that in the case of pentamethylenetetrazole some association occurs in the aqueous phase. For hexamethylene- tetrazole and heptamethylenetetrazole it was assumed that trimers formed in the organic phase, and on the basis of this assumption, an equation was derived from which the trimerization equilibrium constants and the partition co- efficients were calculated. PART I: SPECTROSCOPIC STUDIES OF IONIC SOLVATION PART II: GAS CHROMATOGRAPHIC STUDIES OF CONVULSANT TETRAZOLES BY xi Robert G? Baum A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1976 (aXOLUTIO/V G /o ‘2 .4 3‘ Q? \ «“3 §‘Q’GR‘CA/l/ ’9 A ”76—1916 A BICENTENNIAL THESIS ii ACKNOWLEDGMENTS The author wishes to thank Professor Alexander I. Popov for his guidance, counseling, and friendship through— out this study. Professor Andrew Timnick is acknowledged for his many helpful suggestions as second reader. Gratitude is also extended to the Department of Chemistry, Michigan State University, the National Science Foundation, and the National Institutes of Health of the Department of Health, Education, and Welfare for financial support. Special thanks are given to Dr. Mark Greenberg for many enlightening discussions and for his highly valued friendship, to David DeBrosse for initiating the gas chromatography studies, to Patrick Kelly for interfacing the Raman spectrometer and for providing program Raman, to Eric Roach and Frank Bennis, without whose cooperation the NMR investigations would have been much more difficult, and to Ada Hourdakis and John Hoogerheide for their expendi- ture of time in proofreading this thesis. Appreciation is extended to John Thompson, Drs. Richard Bodner, Paul Handy, Wayne (Duke) DeWitte, Paul Gertenbach, Yves Cahen, and to the present members of the “group" for their friendship and encouragement during our association. I would also like to thank my many friends in the Department of Chemistry for an occasional round of golf, iii several fruitless fishing expeditions, and for many other good times. Deep appreciation is extended to my Mother, who will forever consider me a professional student. Finally, I would like to thank my wife, Linda, for her love, patience, and encouragement while she, too, was pursuing a Ph.D. degree. iv TABLE OF CONTENTS PART I SPECTROSCOPIC STUDIES OF IONIC SOLVATION Chapter I. HISTORICAL . . . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . FAR-INFRARED SPECTROSCOPY. . . . . . . . . . MID-INFRARED AND RAMAN SPECTROSCOPY. . . . . NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY. . . CONCLUSIONS. . . . . . . . . . . . . . . . . II. EXPERIMENTAL . . . . . . . . . . . . . . . . . REAGENTS . . . . . . . . . . . . . . . . SOLVENTS . . . . . . . . . . . . . . . . . WATER ANALYSIS . . . . . . . . . . . . . . . SAMPLE PREPARATION . . . . . . . . . . . . . INSTRUMENTAL MEASUREMENTS. . . . . . . . . Nuclear Magnetic Resonance. . . . . . . Far—infrared Spectra. . . . . . . . . . Laser Raman Spectra . . . . . . . . DATA PROCESSING. . . . . . . . . . . . . . . III. SPECTROSCOPIC STUDIES OF LITHIUM ION SOLVATION IN ACETONE AND ACETONE-NITROMETHANE MIXTURES INTRODUCTION . . . . . . . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . Far- infrared Spectra. . . . . . Nuclear Magnetic Resonance spectra. . . . Laser Raman Spectra . . . . . . . Effect of Complexing Agents . . . . . . AWXE‘ Page 15 18 19 20 20 20 21 21 21 23 24 PART II GAS CHROMATOGRAPHIC STUDIES OF CONVULSANT TETRAZOLES Chapter I. HISTORICAL . . . . . . . . . . . . . . . II. EXPERIMENTAL . . . . . . . . . . . . . . REAGENTS . . . . . . . . . . . . . . . SYNTHESIS OF 8-SEC—BUTYL— PENTAMETHYLENETETRAZOLE. . . . . . . . 4— ec— butylcyclohexanol . . . . . . 4-—__g-buty1cyclohexanone. . . . . . 4-___C—buty1cyclohexanone Oxime. . . 8— ___ cbutylpentamethylenetetrazole. SYNTHESIS OF 8-TERT-BUTYL- PENTAMETHYLENETETRAZOLE. . . . . . . 4-tert-butylcyclohexanone Oxime 8—tert-butylpentamethylenetetrazole MELTING POINTS . . . . . . . . . . . CONSTANT TEMPERATURE BATH. . . . . GAS CHROMATOGRAPHIC MEASUREMENTS ANALYTICAL STUDIES . . . . . . . . Solutions . . . . . . . . . . Chromatographic Response. . . . . . Retention Times . . . . . . . . . DISTRIBUTION STUDIES . . . . . . . . . SOLUBILITY STUDIES . . . . . . CRYOSCOPIC MEASUREMENTS. . . . . III. ANALYTICAL STUDIES OF THE CYCLOPOLY— METHYLENETETRAZOLES AND THEIR MIXTURES BY GAS CHROMATOGRAPHY. . . . . . INTRODUCTION . . . . . . . . . . . . RETENTION TIMES AND SENSITIVITIES. vi Page 58 69 70 76 76 76 76 76 77 78 78 79 79 80 81 82 83 84 Chapter Page RESOLUTION OF A MIXTURE OF TETRAZOLES. . . . . 88 WORKING CURVES . . . . . . . . . . . . . . . . 91 IV. SOLUBILITY AND DISTRIBUTION STUDIES OF THE CYCLOPOLYMETHYLENETETRAZOLES. . . . . . . 96 SOLUBILITY STUDIES . . . . . . . . . . . . . . 97 DISTRIBUTION STUDIES . . . . . . . . . . . . . 98 DISCUSSION OF RESULTS. . . . . . . . . . . . . 107 APPENDICES I. DESCRIPTION OF PROGRAM RAMAN AND THE ACCOMPANYING INTERFACE FOR THE SPEX RAMALOG 4 RAMAN SPECTROMETER . . . . . . . . . . 122 INTERFACE. . . . . . . . . . . . . . . . . . . 122 Triggering Interface. . . . . . . . . . . . 122 ADC Interface . . . . . . . . . . . . . . . 124 OPERATION OF PROGRAM RAMAN . . . . . . . . . . 124 Enabling OS/8 . . . . . . . . . . . . . . . 124 Running Program Raman . . . . . . . . . . . 127 II. DETERMINATION OF PHARMACOLOGICAL PROPERTIES OF 8-TERT—BUTYLPENTAMETHYLENETETRAZOLE AND 8- SEC- -BUTYLPENTAMETHYLENETETRAZOLE AS QUOTED FROM “COMMUNICATIONS RECEIVED FROM PROFESSOR WILLIAM E. STONE . . . . . . . . . . . . 139 LITERATURE CITED . . . . . . . . . . . . . . . . . . . 142 LIST OF TABLES Solvation Band Frequencies of Alkali Metal Ions in Nonaqueous Media. . . . . Variation of the Lithium-7 Resonance as a Function of Mole Fraction of Acetone for Solutions of LiClO4 in Acetone- Nitromethane Mixtures . . . . . . . . . Variation of the Lithium—7 Resonance as a Function of the Acetone/LiClO4 Mole Ratio for Solutions of LiClO4 in Acetone— Nitromethane Mixtures . . . . . . . . . Linewidths of 35Cl Resonance for Solutions of LiClO Values Obtained for the Concentration of Free and Bound Acetone and H Values for Raman Study of LiClO4 in Acetone—Nitro- methane Mixtures. . . . . . . . . . . . Physical and Pharmacological Properties of some Cyclopolymethylenetetrazoles. . Relative Retention Times and Sensitivities for the Cyclopolymethylenetetrazoles. . Solubilities of some Cyclopolymethylene— tetrazoles. . . . . . . . . . . . . . . Distribution Ratios of Pentamethylene— tetrazole, Hexamethylenetetrazole, and viii 4 in Acetone-Nitromethane Mixtures. Page 34 37 4o 51 61 85 97 Table 10 11 12 13 Page Heptamethylenetetrazole at Various Concentrations. . . . . . . . . . . . . . . . . 100 Distribution Ratios for 8-tert-buty1penta— methylenetetrazole between Water and Carbon Tetrachloride. . . . . . . . . . . . . . 103 Distribution Ratios of some Cyclopoly— methylenetetrazoles between Water and Carbon Tetrachloride. . . . . . . . . . . . . . 108 Molecular Weights of some Cyclopoly— methylenetetrazoles in Aqueous Solutions as Determined by Cryoscopic Measurements. . . . 114 Partition Coefficients for Hexamethylene— tetrazole and Heptamethylenetetrazole . . . . . 120 Figure 1 LIST OF FIGURES Page Absorbance of 369-cm_l band XE acetone/ LiClO4 mole ratio. Concentration of acetone = 0.2 M; solvent: nitromethane. . . . . 29 Far-infrared spectra of solutions of LiClO4 in acetone . . . . . . . . . . . . . . . 31 7Li chemical shifts XE mole fraction acetone for solutions of LiClO4 in acetone—nitromethane mixtures. Con— centration of LiClO4 = 0.4 M. . . . . . . . . . 35 7Li chemical shifts YE acetone/LiClO4 mole ratio. Concentration of LiClO4 = 0.4 M; solvent: nitromethane. . . . . . . . . . 38 Band width at half height of 35C1 reson- ance XE acetone/LiC1O4 mole ratio. Con- centration of LiClO4 = 0.4 M; solvent: nitromethane. . . . . . . . . . . . . . . . . . 41 l and 803—cm"l The 789—cm— acetone Raman bands at various acetone/LiClO4 mole ratios. . . . . . . . . . . . . . . . . . . . . 44 Computer—fitted 789~cm_l acetone Raman band for a standard acetone solution in nitro— methane. X means an experimental point, 0 means a calculated point, = means an experimental and calculated point are the same in the resolution of the plot. . . . . . . 47 X Figure 10 ll 12 l3 l4 Page Plot of the area of 789-cm_l acetone Raman band 25 concentration of acetone; solvent: nitromethane. . . . . . . . . . . . . . . . . . 48 1 and 803—cm—l ace— Computer-fitted 789-cm‘ tone Raman bands for a solution of 0.4 M LiClO4 in an acetone-nitromethane mixture. X means an experimental point, 0 means a calculated point, = means an experimental and calculated point are the same in the resolution of the plot. . . . . . . . . . . . . 50 Plot of H Xi the acetone/LiClO4 mole ratio. Concentration of LiClO4 = 0.4 M; solvent: nitromethane . . . . . . . . . . . . . 52 The 935-cm'l C10; Raman band as a function of the PMT:LiClO4 mole ratio. . . . . . . . . . 56 Synthesis of 8-sgg—butylpentamethylene— tetrazole . . . . . . . . . . . . . . . . . . . 75 Plot of the relative retention time gs the molecular weight of the cyclopoly- methylenetetrazoles . . . . . . . . . . . . . . 86 Chromatogram of a mixture of cyclopoly- methylenetetrazoles. Sample: 2.0ul of c3 — cll tetrazoles, each present at a con— centration of 0.0138 g/lO ml of chloroform. Injector: 220°; Column: 210°; Detector: 300°; xi Figure 15 16 17 18 19 20 21 Page He at 60 ml/min; Sensitivity: 128 x 10—11 a.f.s.. . . . . . . . . . . . . . . . . . . . . 90 Plot of average detector response XE concentration of pentamethylenetetrazole in aqueous solution . . . . . . . . . . . . . . 93 Average detector response Ki concen— tration of heptamethylenetetrazole in carbon tetrachloride solution . . . . . . . . . 94 Plot of the distribution ratio XE the con- centration of pentamethylenetetrazole for the water-carbon tetrachloride system . . . . . . . 102 Distribution ratio Ki concentration of hexamethylenetetrazole for the water— carbon tetrachloride system . . . . . . . . . . 104 Distribution ratio Xi concentration of heptamethylenetetrazole for the water- carbon tetrachloride system . . . . . . . . . . 105 Distribution ratio XE concentration of 8- tgrt—butylpentamethylenetetrazole for the water-carbon tetrachloride system . . . . . . . 106 Plot of the distribution ratio K2 the concentration of pentamethylenetetrazole in the aqueous phase for the water—carbon tetrachloride system. . . . . . . . . . . . . . 113 Figure 22 23 Al A2 A3 Page Plot of the distribution ratio gs the square of the concentration of hexamethylenetetrazole in the aqueous phase for the water-carbon tetra— chloride system . . . . . . . . . . . . . . . . 118 Distribution ratio 25 the square of the concentration of heptamethylene— tetrazole in the aqueous phase for the water—carbon tetrachloride system . . . . . . . 119 Triggering interface and 5 volt regulated power supply. . . . . . . . . . . . . . . . . . 125 Analog-digital converter and power supply. . . . . . . . . . . . . . . . . . . . . 126 Flow chart for Program Raman. . . . . . . . . . 131 xiii LIST OF NOMENCLATURE, ABBREVIATIONS AND SYMBOLS Contact Ion Pairs. Pairs of ions, linked electrostatically, but with no covalent bonding between them. Solvent Shared Ion Pairs. Pairs of ions, linked electro— statically by a single, oriented solvent molecule. Solvent Separated Ion Pairs. Pairs of ions, linked electro— statically but separated by more than one solvent molecule. MeZCO: Acetone PMT: Pentamethylenetetrazole C6CMT: Hexamethylenetetrazole C7CMT: Heptamethylenetetrazole 8-Egrt—buty1 PMT: 8-tg£t-buty1pentamethylenetetrazole B-EEE—butyl PMT: 8-seg-buty1pentamethy1enetetrazole T: One Tesla = 10 kilogauss ATf: Freezing point depression in °C. f________________________________________________________—________TET__T“_———i" PART I SPECTROSCOPIC STUDIES OF IONIC SOLVATION CHAPTER I HISTORICAL INTRODUCTION Most chemical reactions whether in research labora— tories, industry, or biological systems occur in solutions. The association of solvent molecules with metal ions in solutions has been an important field of research for many years, and a wide variety of experimental techniques have been used to study ion—ion, ion—solvent, and solvent— solvent interactions. However, we still have only a rudimentary concept of the structure of electrolyte solu- tions. Classical techniques such as electrochemical measure— ments and the study of colligative properties of solutions have been widely used in studies of electrolyte solutions. These methods, however, measure bulk solution properties and give little information about the chemical nature of the species present in solution. For example, it is very difficult to distinguish between contact and solvent separated ion pairs by the above techniques. Similarly, attempts to determine ionic solvation numbers often give contradictory results. In many cases no clear-cut distinc— tion can be made between the inner and outer solvation spheres and, consequently, for a given ion, vastly dif— ferent solvation numbers can be obtained with different experimental techniques (1). Within the last decade spectroscopic techniques such as mid— and far—infrared, Raman, and nuclear magnetic resonance have been very useful probes for the elucida— tion of the structure of electrolyte solutions and of the chemical species present in them. FAR-INFRARED SPECTROSCOPY In 1965 Evans and Lo (2) studied the far-infrared spectra of tetrabutyl— and tetrapentylammonium halides in benzene solution. They observed a band in the lOO—cm-l spectral region which could not be assigned to a vibration— al mode of either the solvent or the solute. Since the band position was dependent on the mass of both the cation and anion, the authors assumed that it was due to a cation—anion ion pair vibration. This constituted the first report of an ionic vibration in solution. Shortly thereafter, Edgell and co—workers (3,4) ob— served far-infrared bands due to the motion of the alkali cation in tetrahydrofuran solutions of lithium, sodium, and potassium tetracarbonylcobaltate and pentacarbonyl- manganate. Upon extending these studies to other solvents (dimethylsulfoxide, pyridine, and piperidine), the authors observed the band position to be a function of the cation and the solvent. Popov and co—workers (5—14) extended these far—infrared studies to include several nonaqueous solvents. They found that in highly solvating solvents such as dimethyl— sulfoxide (5,6), the frequencies of the bands are strongly dependent on the nature of the cation but are completely independent of the anion. They concluded that these bands were due to the alkali metal ion vibrating in a solvent cage. Thus these bands were named "solvation bands". However, in solvents with very low solvating abilities such as tetrahydrofuran (3,4), some anion de- pendence is observed. It was postulated that this de— pendence is due to a change in the nature of the solvent cage around the cation. In these cases a counter—ion re- places a solvent molecule in the inner solvation shell and forms a solvated contact ion pair. Thus the cation is vibrating in a cage composed of solvent molecules and a counter—ion. While studying the far-infrared spectra of sodium tetrabutylaluminate in cyclohexane solutions, Tsatsas and Risen (15) noted two solvation bands at 195 and 160 cm_l. In tetrahydrofuran solutions, however, only one band, at 195 cm-1, was observed. In addition, a Raman band at 202 cm_1 was seen for the cyclohexane solutions. However, Edgell at El. (3,4) had previously shown that the far— infrared solvation bands are Raman inactive, which is in— dicative of the electrostatic nature of the ion—solvent or ion-ion interaction. Thus in cyclohexane solutions of sodium tetrabutylaluminate, the ion—solvent interaction 1 . . . . . Vibration possesses a Slgnlfl- responsible for the 202—cm— cant degree of covalency. Wong gt al. (9) investigated solutions of lithium perchlorate in acetone and acetone—nitromethane binary solvent mixtures. They observed changes in the far-infra— 1 red region of the spectrum where the 390-cm— acetone band (C—C—C deformation) was split upon the addition of the salt, and a new band appeared at 369 cm_l. This new band was assigned to be due to a vibration of acetone complexed to the lithium ion. Regis and Corset (16), however, re— cently disagreed with this conclusion and stated that l was due to the lithium ion vibrat— the band at 369 cm— ing in a nitromethane solvent cage. Erlich at El- (14) studied the variation in the fre- quency of the sodium solvation band in dimethylsulfoxide- pyridine mixtures. As the solvent composition was changed, the frequency of the solvation band progressed gradually from the frequency characteristic of one solvent to that characteristic of the other. They observed a strong preferential solvation of the sodium ion by dimethylsul— foxide. Recently, Barker and Yarwood (17) extended the work of Evans and Lo (2). In studying the far—infrared spectra for benzene solutions of tetrabutylammonium chloride, they noticed an asymmetry to the low—frequency side of l the 115-cm_ band which suggested the presence of unre— solved bands. They showed that a second band was present at N75 cm—l, and they attributed this band to a perturbed "collisional" or lattice band of the benzene molecule. This implies that the benzene solvent molecules are sol— vating the aggregate that gives rise to the band at 115 cm—1. Far-infrared spectroscopy has also been used in Studying complexation reactions of alkali metal ions. Risen and co—workers (18) observed far—infrared bands for lithium, sodium, and cesium ions in ethylene-meth— acrylate copolymers. They also observed far-infrared bands for alkali metal ion complexes of cyclic polyether compounds in dimethylsulfoxide and pyridine solutions (19). More recently, Cahen and Popov (20) observed the far-infrared spectra of sodium and lithium cryptates in several nonaqueous solvents. The spectra were characterized by a broad band whose frequency was independent of the solvent and the anion. The band was assigned to the vibra— tion of the cation in the cryptand cavity. The frequencies of the alkali metal ion solvation bands in several solvents are presented in Table l. MID-INFRARED AND RAMAN SPECTROSCOPY Day and co—workers (21,22) observed that some infrared bands of tetrahydrofuran are split by the sodium ion, giving rise to new bands characteristic of tetrahydrofuran bound to the sodium ion. From the band intensity measure— ments, they determined that the sodium ion is solvated by four tetrahydrofuran molecules and calculated stability constants for the stepwise complexation of the sodium ion by tetrahydrofuran. A similar technique was used by Taylor and Kuntz (23) in their study of anion solvation ma NHH mHH va vwa mma oov mpmconumo mqwamaoum NH mmm wchHummouoHnulm NH NON mcHOHHNmHNnumsHo-e.m NH NON mcHOHHNmHsnumeHoue.N NH OON wcHOHuNmHsrumz-m NH oom omH oam wchHuhmenuoznw HH oom omH omv oCHCHHSm OH OON OHo< oHuw64 O OOH OOH mNe mcopmo< N OHO muoeHHouuNm-N-HNaH>-H O.N OOH OOH NON OON OON wsoeHHouHNm-N-HNrumz-H 8 N mOH OHN NON OOO chOHHouuleN O NmH ONN ONN mNO mOHxOHHSmHNusto O mNH mmH NNN ONN ONe mOonNHsmNdoumHo 0.0.m OHH mNH mmH eHN OON ONe mOonuHsmHNruweHo v omH mchHuwm e mmH wCHUHHomHm O.e OON mNO OO-wOHx0uH5mHNeHmeHo mH.mH omH OOH NOO cmuzmogescmuowe wocmuwmmm +mo +Qm +M Nmz +wz +HH mwcm>aom AHIEUV meocosmem pcmm :oHum>Hom .meoz msomswmcoz cH mCOH Hobo: HmeH¢ mo moHocmsonm pcmm coflum>fiom .H magma by phenol. The solvation of the lithium ion by dimethylforma- mide (DMF) was investigated by Lassigne and Baine (24) by infrared and nuclear magnetic resonance (NMR) spectros- copy. They monitored the infrared carbonyl band of DMF at 1686 cm_l. As lithium perchlorate was added to the solution, a new band appeared at 1670 cm—1, which was in- dicative of DMF bound to the lithium ion. The solvation number of lithium by dimethylformamide was found by NMR to be four. Wong gt El: (9) studied the vibrations of the per- chlorate ion and of acetone in solutions of lithium per- chlorate in acetone—nitromethane mixtures. It was noted that the 935-cm-l Raman band (v1 symmetric stretch) of the perchlorate ion remained narrow and at constant frequency in solutions with acetone/lithium mole ratios of :4, but that it broadened and shifted to higher frequency as the acetone/LiClO4 mole ratio became less than four. It was concluded from these data that the inner solvation shell of Li+ contained four acetone molecules. Similar results were reported by Handy and Popov (12) in their study of the solvation of the lithium ion by 4-methyl-pyridine. In a series of studies, Edgell's group (25—27) used the carbonyl stretching frequency of the tetracarbonyl- cobaltate anion as a probe of the environment of this anion. l The l900-cm— (>«D stretch was monitored for solutions of NaCo(CO)4 in several solvents as a function of temperature and of salt concentration. In dimethylsul- foxide, dimethylformamide, nitromethane, hexamethylphos- phoramide, acetonitrile, and pyridine the band was quite symmetrical which indicates that only solvent molecules are near—neighbors of the Co(CO)2 ion in these solutions. However, in piperidine, tetrahydrofuran, and dimethoxy— ethane solutions, additional bands were observed at the high— and low—frequency side of the main band. The be- havior of the two new bands indicates an increasing asym- metrical environment about the anion resulting from contact ion pairing. They observed the spectra of tetrahydrofuran solutions of NaCo(CO)4 at various temperatures. This en- abled them to resolve the complex spectra into four band components, which indicated the presence of solvent sepa- rated and contact ion pairs. Edgell gg gi. (28) recently investigated the infrared spectrum of thallium tetracar- bonylcobaltate in seven solvents. Only a single ionic environment was found in dimethylformamide, dichloro- methane, and dimethylsulfoxide solutions. Several kinds of environments were found in tetrahydrofuran, acetoni- trile, and nitromethane solutions which resulted from solvent surrounded ions, contact ion pairs, and triple ions. Borucka and Kecki (29—31) studied the infrared spectra of several electrolytes in acetone solutions. They noted the splitting of the v and vibra— c-c-c Vc=o tional bands of acetone and correlated the frequencies ll of the new components with the charge density of the cations. They also observed the influence of various metal perchlorates and of lithium and zinc halides on the integral intensities of the acetone vibrational bands. In addition, the band frequency changes were related to the electronic structure of acetone molecules complexed with the cations and anions. In a recent series of papers, Perelygin and Klimchuk (32-36) examined infrared spectra of alkali and alkaline earth metal salts in nonaqueous solutions. For solutions of sodium, lithium, and magnesium perchlorates in aceto- nitrile (32), they observed perchlorate vibrational bands in the llOO—cm-l spectral region which were due to free ions and contact ion pairs. They also noted changes in l (CEN stretch). the acetonitrile Vibration at 2254 cm- This band is shifted by 10, 21, and 36 cm"1 respectively as a consequence of bonding of the acetonitrile molecule with the sodium, lithium, and magnesium cations. From measurements of the band intensities they calculated the association constants of the ions and the coordination numbers of the cations. For Na+, Li+, and Mg2+ ions the corresponding solvation numbers are four, four, and six. They also investigated the solvation numbers as a function of temperature and found that between —40 and 60°C they did not vary from the above values (33). For acetonitrile solutions of sodium and lithium iodides, they determined the ionic association constants and also found that the 12 solvent sheath of the iodide ion contained eight aceto— nitrile molecules (34). Similar studies were carried out on acetone solutions of sodium, lithium, and magnesium perchlorates (35,36). The solvation numbers of the cat- ions were calculated to be four for Na+ and Li+ and six for Mgz+, as for the solutions in acetonitrile. Raman spectroscopy is being used increasingly to investigate ionic association and solvation in solutions. The vibrational modes of polyatomic anions such as nitrate, perchlorate, and sulfate are very sensitive to their en- vironment and can be used as probes in the studies of ionic interactions. Peleg (37) examined the magnesium nitrate-water system by Raman spectroscopy. The vibrations of the nitrate ion were observed over the range from very dilute solution to the anhydrous molten salt. The results in- dicated that the interactions in the system varied as the composition varied. In very dilute solutions both ions are completely hydrated, and the nitrate ion is per— turbed by the water molecules. As the water content is lowered, the polarization power of the magnesium ion be- gins to affect the nitrate ion, but no contact ion pair- ing occurs until the water content is reduced below six moles of water per mole of salt. Upon further decrease of the water content both contact and solvent separated ion pairs exist in solution. From these results it was suggested that the solvation number of the magnesium l3 cation by water is six. The magnesium nitrate-water system was also studied by Chang and Irish (38) by infrared and Raman spectros- copy. By very careful computer resolution of the spec- tral bands, they showed that as the water content de— creases, solvent separated ion pairs give way to contact ion pairs in which the nitrate ion is bound to the Mg2+ ion in monodentate fashion. On further reduction of the water content, the nitrate becomes bound to the M92+ ion in a bidentate fashion. Solutions of silver nitrate in acetonitrile and water were recently investigated by Chang and Irish by infrared and Raman spectroscopy. For acetonitrile solutions (39) they concluded that the nitrate ion exists in three differ— ent environments. At salt concentrations of 4 M or less, both free nitrate ions and ion pairs are present in solu— tion. When the salt concentration is greater than 4 M, they noted the formation of multiple ion aggregates. The association constant for the Ag+NOS ion pair was calculated, and from a plot of the average number of acetonitrile mole- cules bound to the silver ion, 3, versus the AgNO3/aceto— nitrile mole ratio, they determined the solvation number of Ag+ by acetonitrile to be four. Aqueous silver nitrate solutions (40) were described in terms of an equilibrium between free ions and ion pairs. The ion pair association 1 was obtained from monitoring both the 1 constant of 0.1 M— 717—cm_l and 1047—cm_ nitrate vibrations. 14 The AgNO3-—acetonitrile system was investigated further by Janz and Mfiller (41). They examined ion pair— ing by careful and precise Raman measurements and extended these studies to very dilute solutions, so as to overlap the concentration range over which the Fuoss—Onsager con- ductance theory applies. The value of the ion pair associ— ation constant calculated from the Raman data is 84tl4; from the Fuoss—Onsager conductance theory, the value ob— tained is 70.3il.2, which shows good agreement between the two experimental techniques. The complexation of the cadmium ion by nitrite ion in aqueous solutions was studied by Irish and Thorpe (42). Upon addition of Cd(ClO4)2 to aqueous solutions of NaNOz, l a new Raman band appears at 861 cm_ and grows in inten- sity as the Cd2+/NOS mole ratio increases. At the same time, the intensity of the nitrite band at 817 cm—1 de— creases. The free nitrite concentration was monitored and the average ligand number was evaluated. Four suc— cessive formation constants were determined for the species Cd(N02)+, Cd(NO Cd(N02)', and Cd(NO It 2)2' 2)i_° was also determined that the chelation occurs through the two oxygen atoms. Plowman and Lagowski (43) examined Raman spectra of solutions of alkaline earth and alkali metal perchlor— ates and nitrates in liquid ammonia. They observed low- frequency bands of Li+, Na+, Mg2+, Ca2+, Sr2+, and Ba2+ at 241, 194, 328, 266, 243, and 215 cm’l, respectively. 15 These bands were assigned to the symmetric stretching mode of the solvated cation, which presumably originate from interaction of the first solvation shell. The results presented in this thesis are, in general, applications of the aforementioned studies. A more exten— sive historical discussion of solvation studies by vibra- tional spectroscopy can be found in the doctoral disserta— tions of B. W. Maxey (44), J. L. Wuepper (45), M. K. Wong (46), and P. R. Handy (47). NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY During the past few years numerous publications have appeared which deal with solvation studies by proton nuclear magnetic resonance (NMR) spectroscopy. In electro— lyte solutions the solvent molecules can exist in several environments. These environments may be divided into bulk solvent regions, where solvent molecules are ef— fectively out of range of ionic influence, secondary solvation regions, and primary solvation regions. If the exchange of molecules between all of these environments were very slow, a number of resonance lines would be ex- pected in the NMR spectrum of the solvent nuclei that correspond to the different environments. Generally, however, the exchange is quite rapid. Thus, the separate resonance signals for each environment are population averaged to a single line whose shift from the pure sol- vent resonance signal reflects the average effect of 16 the different environments. Two proton nuclear magnetic resonance methods have been widely used for the determination of ionic solvation numbers. The first involves the cooling of the solution in order that the proton exchange be slowed to an extent that separate signals can be observed for the coordinated and bulk solvent molecules. This method has been exten— sively employed by Fratiello and co-workers (48—50) in the determination of hydration numbers of several metal ions. In the second method, the chemical shift of the solvent protons is monitored as the solvent/salt mole ratio is varied, and the results are plotted. Often a distinct break in the resulting curve is observed that indicates the solvation number. This technique was used by Schaschel and Day (51) to complement their results obtained by in— frared spectroscopy (21,22). In studying the solvation of the sodium ion by tetrahydrofuran (THF), they monitored the chemical shift of the THF protons as a function of the THF/NaAlBu4 mole ratio. From the results the authors concluded that Na+ was solvated by four THF molecules. Although proton nuclear magnetic resonance has been extensively used to investigate electrolyte solutions, the protons are usually several atoms removed from the actual site of interaction and, consequently, the chemical shifts are only weakly affected by the solvation. Thus, ambiguous results are often reported. It is obvious that better information can be obtained by observing the IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII[:::________________—_____————————————T:;7 , l7 resonance of the solvated species. Alkali metal and halogen NMR have been used as very sensitive probes in the elucidation of the structure of alkali salt solutions in nonaqueous solvents. Several ex— tensive historical reviews of non—proton NMR have recently been written. A comprehensive discussion of 23Na and 7Li NMR can be found in the doctoral dissertations of M. S. Greenberg (52) and Y. M. Cahen (53), respectively. The use of alkali metal NMR (especially 133Cs ) and of halogen NMR in solvation studies was recently reviewed by DeWitte (54). Therefore, the remainder of this discussion will be devoted to a few recent examples. Greenberg and Popov (55) published results obtained from 23Na NMR studies of preferential solvation in non— aqueous mixed solvents. Generally, these studies reflected the relative donicity of each solvent in a given solvent pair, where the solvent of higher donicity was preferen- tially contained in the inner solvation shell of the Na+ ion. Cahen gg g1. (56) examined the solvation and ion pair formation of several lithium salts in nonaqueous solu- tions by using 7Li and 35Cl NMR. Formation of contact ion pairs in tetrahydrofuran, nitromethane, and tetramethyl- guanidine was particularly evident as considerable broad- ening of the 35C1 resonance of the perchlorate ion was observed. Similar results were recently obtained by Berman and Stengle (57) in their studies of metal perchlorate 18 solutions in nonaqueous solvents. Very recently the solvation of hexafluorophosphate l9 salts was investigated by DeWitte and Popov (58) by F 23 and Na NMR measurements. They also reported on the solva- tion and ionic association of several cesium salts in non- 133Cs NMR (59). aqueous solutions by The use of alkali metal NMR has yielded considerable information as to the nature of alkali metal ion complexes. Cahen gg gl. (60,61) determined the formation constants and studied the complexation reaction kinetics for various lithium—cryptand complexes, while Mei gg gl. (62) investi— gated complexes of Cs+ with cyclic polyether compounds 133 ("crowns") by Cs NMR. CONCLUSIONS It thus appears that spectroscopic techniques can be very useful in the elucidation of the structure of electro— lyte solutions. Now, with the advent of laser excitation Raman spectroscopy and Fourier transform infrared and nuclear magnetic resonance spectroscopy, very dilute solu— tions can be investigated. This makes possible more direct comparisons of results obtained by spectroscopic methods with those obtained by classical techniques. CHAPTER II EXPERIMENTAL l9 REAGENTS Lithium perchlorate (Fisher) was dried at 190°C for several days. The water content was found to be 0.2% by weight. Lithium iodide (K & K Laboratories) was prepared as previously described (56). Solutions of lithium tri— iodide were prepared by the addition of equimolar amounts of iodine to lithium iodide solutions. Reagent grade iodine (Baker) was used without further purification. Penta— methylenetetrazole (Aldrich) was recrystallized from diethyl ether and dried i2 vacuo. SOLVENTS Reagent grade acetone (Baker) was refluxed over Drie— rite and then fractionally distilled. The acetone was further dried over freshly activated 5A Linde molecular sieves and stored in a dry box under nitrogen atmosphere. Acetone—d6 (Aldrich, 99+%) was dried over molecular sieves and stored in a dry box. Spectroscopic grade nitromethane was fractionally distilled and dried over molecular sieves for 24 hours. Water content was found to be <50 ppm. The molecular sieves used were activated by heating them at 500°C under dry argon for 12 hours. WATER ANALYSIS Analyses for water in salts and solvents, where possible, were carried out with an Aquatest II (Photovolt Corp.) 20 21 automatic Karl Fischer titrator. SAMPLE PREPARATION Generally, solutions of lithium salts were prepared by weighing out the desired amount of salt into a 5 or 10 ml volumetric flask, transferring the flask to the dry box, and then diluting to the mark with solvent. The acetone-nitromethane mixed solvent solutions were prepared by taring a small snap-cap vial, adding the desired volume of acetone, weighing, adding the desired amount of nitromethane, and weighing again. From these weights, the solvent composition was determined. INSTRUMENTAL MEASUREMENTS Nuclear Magnetic Resonance Lithium—7 nuclear magnetic resonance measurements were made on a Varian Associates DA—60 spectrometer at a field of 1.4092 T and a frequency of 23.287 MHz. The spectrometer was frequency locked to an appropriate reference solution. The 7 Li chemical shifts were measured against an aqueous 4.0 M LiClO4 solution contained in a 1 mm melting point capillary and centered in the Wilmad 506-PP 5 mm OD polished NMR sample tube by Delrin spacers; however when the chemical shifts were so small that the sample was masked by the ref- erence, a secondary reference of 5.0 M LiClO4 in methanol was used. In the latter case, the shifts were corrected 22 to the 4.0 M aqueous LiClO4 reference solution. A positive shift from the reference is assumed to be upfield. The chemical shifts reported are corrected for dif— ferences in bulk diamagnetic susceptibility between the sample and the reference according to the equation 2r ref _ Xiample) (1) 6corr = Gobs +'3_(Xv ref and xsample v v are the volume susceptibilities in which X of the reference and sample solutions, respectively. Gobs and scorr are the observed and the corrected chemical shifts. Published values of magnetic susceptibilities for the sol— vents (63) were used to calculate acorr. Chlorine-35 NMR measurements were also performed on the DA-60 spectrometer at a field of 1.0378 T and a fre— quency of 4.33 MHz. The spectra were obtained by using the modulation technique previously described (56). All measurements were made at room temperature (25°C) and cylind— rical nonspinning sample tubes of about 15 mm diameter were used. The spectra were recorded in the dispersion mode and linewidths were determined with an estimated accuracy of 110% as an average of two to four measurements. Far—infrared Spectra The far-infrared spectra were obtained with a Digilab FTS—l6 spectrometer. The FTS—l6 is essentially a rapid— scan Michelson interferometer operated under computer Pg. 23 control. The theory and operation of this instrument have been previously described (47). Most of the spectra were obtained by using the 3- or 6—um mylar beam splitters which cover the ranges of 600-150 and 425—100 cm—l, respectively. Most of the spectra were obtained at nominal resolutions of 1 either 2 or 4 cm_ , which give a data point every 1 or 2 cm‘l, respectively. The instrument was operated in the single beam mode. The reference spectrum was stored in the computer memory and subtracted from the solution spectra. Standard demountable cells (Barnes Engineering Co.) were used with 2-mm polyethylene discs, and the path length was maintained at 0.1 or 0.2 mm. All spectra were smoothed by using the 9—point smoothing routine developed by P. R. Handy (47). ' Laser Raman Spectra Raman spectra were obtained on a Spex Ramalog 4 Laser— Raman spectrometer equipped with a model 1401 double mono— chromator. The 5145 A line of a Spectra Physics model 164 argon ion laser was used as the excitation source, and the data were obtained in the pulse counting mode. Samples were injected into 1.6—1.8 X 90 mm melting point capillary tubes and sealed. In most cases the instrument was inter- faced with a Digital Equipment Corp. PDP—8/E lab minicom- puter to obtain digitized spectra. (This program is listed and its application described in Appendix I.) The digi— tized spectra were punched onto paper tape and then transfered 24 to cards for computer analysis. DATA PROCESSING Extensive use of the CDC—6500 computer was made to evaluate the Raman spectral data. The Fortran IV program KINFIT (64) was employed to fit and resolve the spectral bands. The application of this program was described by M. S. Greenberg (52). CHAPTER III SPECTROSCOPIC STUDIES OF LITHIUM ION SOLVATION IN ACETONE AND ACETONE-NITROMETHANE MIXTURES 25 INTRODUCTION Previous studies in our laboratories (9,56,65) and elsewhere (21,23,27,43) have shown that nuclear magnetic resonance as well as infrared and Raman vibrational spec— troscopy are very useful probes for the elucidation of the structure of electrolyte solutions and of the species present in them. These techniques are primarily sensitive to the nearest-neighbor interactions and have been used by several investigators for the determination of primary solvation numbers of ions. In this study far—infrared, Raman, and nuclear mag— netic resonance spectroscopic techniques were employed in the investigation of lithium perchlorate solutions in ace- tone and acetone—nitromethane mixtures. This work was undertaken to study quantitatively the solvation of the lithium ion by acetone and to investigate the solvating ability of nitromethane as compared to that of acetone to determine if the results agree with those reported by Regis and Corset (16). In addition, the influence of a weak complexing agent, pentamethylenetetrazole, on the Li+ClO4 ion pair formation was investigated. RESULTS AND DISCUSSION Far—infrared Spectra In performing a quantitative solvation study, it is very desirable to control the concentration of the 26 27 solvating agent. This procedure, however, requires the use of an inert solvent as diluent. In previous work (9), nitromethane was used as the inert solvent. If, however, the interpretation of Regis and Corset (16) is correct 1 infrared band, observed for solutions (i.e., the 369-cm— of LiClO4 in acetone—nitromethane mixtures, is due to Li+ solvated by nitromethane), then nitromethane competes with acetone for sites in the lithium ion solvation sphere and thus would be a very poor choice for the diluting sol- l far—infrared vent. To determine whether the 369-cm- band is due to a complexed acetone vibration or to a lith- ium-nitromethane vibration, the nitromethane—acetone- lithium ion system has been studied in some detail. Previously, Wong g: gi. (9) monitored the intensity 1 Li+—acetone solvation band as a function of the 425-cm_ of the lithium perchlorate concentration in nitromethane solutions which were 1.5 M in acetone. A linear Beer's law plot of intensity versus concentration was obtained for solutions which were 10.4 M LiClO4. At higher concen- trations, however, although the plot remained linear, the slope was considerably different. It was concluded from these data that below 4:1 acetone/Li+ mole ratio a new absorbing species was formed, which indicated either a possible change in the solvation number of Li+ or a re— placement of an acetone molecule in the solvation shell by the perchlorate ion. In this investigation the intensity of the 369—cm—l 28 band was monitored as a function of the acetone/Li+ mole ratio for solutions of 0.2 M LiClO4 in nitromethane in which the acetone concentration was varied. A near linear relationship was observed (Figure l) for solutions in which the acetone/Li+ mole ratio was :4. Beyond this point, however, the curve began to deviate from linearity. If it is assumed that the 369—cm_1 band is due to complexed acetone, the data indicate that as the concentration of acetone increases, the concentration of complexed acetone increases proportionally; but when the acetone/Li+ mole 1 band begins ratio reaches four, the intensity of the 369-cm— to level off since the excess acetone does not solvate the lithium ion any further. If the 369—cm_l band were due to a lithium—nitromethane vibration as suggested by Regis and Corset (16), its intensity should be fairly high when no acetone is present, but should decrease as acetone is added to the system. Thus, these data seem to indicate that the 369—cm—l band is due to complexed acetone. To confirm this assignment further, far—infrared spectra were obtained for solutions of LiClO4 in nitrometh— ane which were 1.5 M in acetone-d6. In no case was there evidence of a band at 369 cm_l. In addition, the far—infra— red spectra of lithium perclorate solutions were obtained in pure acetone (Figure 2). It is seen that as the con— centration of LiClO4 increases, the intensity of the 390- cm"1 band (C—C—C deformation) decreases, while at the same time the intensity of the 369-cm_l band increases. Thus, 29 .oconquOHuHcluucQ/How um m.o u ocouwom wo .OHDMH mHOE HNOHU..H.H\mEoO.mom m> pawn Eoumom wo wocmnuomod .H 923.3 COHHMHHCwUCOU HI 0:5. 302 II: 00 0—2 ON mp 0p 3 NP 0— m 0 V N o H H H H 1 J H H H H O 1 No.0 .. v0.0 C . 8O sn< 1 mod 30 Figure 2. Far—infrared spectra of solutions of LiClO4 in acetone. INTENSWY V 0 0.6 M 1.0! \\\\\\\\\\///rV/\\\\////// 213M \W 3M + i : : 4 500 450 400 350 300 cm‘ Figure 2 32 these data give additional strong indications that lithium perchlorate causes the splitting of the 390—cm_l band and that Regis and Corset (16) were in error when they assigned the 369—cm_l band observed in acetone—nitromethane mixtures to lithium—nitromethane vibrations. Far—infrared spectra were also obtained for solutions of lithium perchlorate and lithium triiodide in nitro- methane. In the case of the LiClO4 solution, a broad band was observed at 368 cm-1, which confirmed the results of Regis and Corset (16). This band must be due to the Li+ ion vibration in a nitromethane solvent cage. Consequently, it appears that while the band at 368 cm_1 in LiClO4 solu— tions in pggg nitromethane is indeed due to the Li+—nitro— methane vibration, the 369—cm_l band in acetone-nitro— methane mixtures is a displaced acetone band. For solutions of LiI3 in nitromethane, however, the solvation band occurs at 340 cm_l. Thus, the frequency of the Li+ solvation band in nitromethane depends on the anion. It seems reasonable to assume that in these cases some ion pair formation may take place. Therefore, the solvation shell must incorporate anions as well as solvent molecules, thus making the frequency of the Li+ solvation bands anion dependent. To compare the solvating ability of acetone with that of nitromethane, far—infrared spectra were obtained for nitromethane solutions of lithium triiodide to which varying amounts of acetone were added. As the concentration 33 of acetone increased, the 340-cm_1 band was replaced by two new bands at 425 cm—1 1 and 369 cm— which were attributed to the Li+—acetone solvation band and the complexed ace— tone band, respectively. These data again indicate that acetone solvates lithium ion much more strongly than does nitromethane. Nuclear Magnetic Resonance Spectra To establish further the relative inertness of nitro— methane in cation solvation, 7Li NMR studies were carried out on solutions of lithium perchlorate in acetone-nitro- methane mixtures. Two studies were performed. In the first study the 7 Li chemical shift was monitored as a function of mole fraction of acetone for solutions which were 0.4 M LiClO4 in various acetone—nitromethane binary mixtures. The data are presented in Table 2. From the results, it is seen (Figure 3) that the chemical shift progresses smoothly and quite rapidly from that in neat nitromethane (+0.76 ppm) to that in neat acetone (—l.05 ppm). The fact that a curve is obtained and Egg a straight line is indicative of preferential solvation by one of the sol- vents. In this case, the data indicate that acetone sol- vates Li+ much more strongly than does nitromethane, since the limiting shift of LiClO4 in acetone is reached at a low mole fraction of acetone. Since the most dramatic change in the chemical shift occurs between acetone mole fractions of 0 to 0.3, it was decided to extend this study, 34 Table 2. Variation of the Lithium-7 Resonance as a Function of Mole Fraction of Acetone for Solutions of LiClO4 in Acetone-Nitromethane Mixtures. Mole Fraction Acetone Appm 0.0000 0.76 0.0365 0.00 0.1032 -O.59 0.1486 —0.76 0.1955 —0.87 0.2514 —0.95 0.3286 —1.01 0.4019 -1.02 0.6007 -1.07 0.7967 —l.05 1.0000 -l.05 .fl v.0 n «OHUHH mo coHumHDcwocou .moHDDxHE wmmzuoEOHuHCIocoumom CH voHUHH mo mcoHpsHOm How ocoumom mo :oHuomHm mHoE m> mHMHLm HmoHEogo HAN .m ousmHH m20hm0< 20:043.“. wHOE OH O.O .3 O 9O 3 to no NO 3 O . H H H H n H H H H H nfiul 35 2.: d 36 by focusing on the region where the most sudden changes were occurring. In this investigation, the 7Li chemical shift was again monitored, but this time as a function of the acetone/Li+ mole ratio. The data are listed in Table 3. As noted above, the shift again smoothly progresses from that characteristic of lithium perchlorate in nitro- methane to the limiting value of —l.05 ppm in neat ace— tone (Figure 4); however, this limiting shift is obtained in a solution in which the acetone/Li+ mole ratio is 15:1. Therefore, at this mole ratio nitromethane no longer con— tributes to the primary solvation shell of the Li+ ion, which again indicates that the lithium ion is preferen— tially solvated by acetone. It is interesting to note that the isosolvation point (the composition at which the chemical shift is midway between two limiting values) occurs at an acetone/Li+ mole ratio of about 2:1. At the isosolvation point, it has been postulated that there is equal competition be- tween the two solvents for sites in the cation solvation shell. In this case, the data may be indicative that when the primary solvation shell of Li+ is half-filled, it contains two molecules of acetone. Thus we might con— clude that when the Li+ solvation shell is completely filled by acetone, four solvent molecules are present. Additional information was obtained by studying the 35Cl nuclear magnetic resonance in LiClO4 solutions. Gen- 35 erally, the width at half height of the Cl resonance in 37 Table 3. Variation of the Lithium—7 Resonance as a Function of the Acetone/LiClO4 Mole Ratio for Solutions of LiClO4 in Acetone-Nitromethane Mixtures. Acetone/Li+ Mole Ratio Appm l 0.17 2 -0.11 3 -O.28 4 -O.43 5 -0.56 6 -0.66 7 -0.72 8 -0.80 9 —O.83 10 -O.86 15 -O.89 20 -l.05 38 .ocmzuoEOHuH: Humm>H0m “w v.c u vOHUHH mo cofluwuucoocou .oHumm oHOE OOHUHH\o:oumow m> mDMHSm HMUHEono HAN .v ouomHm L.— O:.H0m Hm v.0 n vOHUHH mo :oHpmuucoocoo .oHumH . o oHOE vOHUHH\wcouwom mM mocmCOme HUmm m0 uanon meg on SDUH3 comm m mus Hm 0.5: 30.2 .s wZObwU< a) 3.: 3. 42 system is deficient in acetone, the perchlorate ion fills the vacant position in the solvation sphere of the lithium ion in preference to nitromethane. This conclusion is in agreement with other work (65,66) which has shown that the contact ion pair equilibrium strongly depends on the donor ability of the solvent as well as on the bulk di- electric constant of the medium. Although nitromethane has a high dielectric constant of 35.9, its donor ability is quite low, as shown by the Gutmann donor number of 2.7 (67). Acetone, on the other hand, has a lower dielectric constant, 20.7, but its donor number is 17.0 (67). Although it cannot be stated that nitromethane is devoid of all solvating ability, its relative inertness shows that it is a poor competitor of acetone for positions in the inner solvation shell of the lithium ion. Laser Raman Spectra In the investigation of solutions of LiClO4 in acetone- nitromethane mixtures by Raman spectroscOpyr a change in the appearance of the 789-—cm"l acetone band (methyl de- formation) with the changing composition of the solvents was observed by Wong gg gl. (9). As the concentration of lithium perchlorate was increased, a new band, characteris- tic of the acetone molecule bound to the lithium ion, ap— peared at 803 cm_l. The purpose of this study was to monitor carefully the behavior of these two bands to obtain some quantitative information on the strength of the 43 lithium—acetone interaction. 1 1 As can be seen in Figure 6, the 789—cm- and 803-cm— bands overlapped to an extent which could not be ignored if this investigation was to yield quantitative results. Therefore, digitized spectra were obtained and were analyzed by computer in order to resolve the overlapping spectral bands. The equation used for fitting the experimental data is the Lorentz—Gaussian product function described by Irish 2: sl- (68): I = Io{expE m mo uon +_._ 0:5. 3oz mzofio< op n H Hi,IH H — F——-‘-O .OH stOHc 53 In his fundamental treatment of stepwise equilibria that give rise to species MAl' MA2,..., MAn with correspond- ing equilibrium constants of K1’ K2,...,Kn, Bjerrum (69) has shown that in a solution when H = n - 8, equal concen- trations of MAn_l5 and MAn must exist. Then it follows that, Kn = l/[(A)flfign_% (4) where (A)f is the concentration of free ligand at Hen—%. Thus, the value of K1 is given by [l/(A)f]fié%, K2 is given by [l/(A)f]fig3/2, etc. By using this technique estimates of the stepwise stability constants for the complexation of the lithium ion by acetone have been made. The values obtained are: Kl = 19.1, K2 = 2.5, K3 = 1.3, and K4 = 0.6. It should be emphasized that due to the uncertainty inherent in this method, the calculated constants are probably only an indication of order of magnitude. Effect of Complexing Agents In previous work (9) it was noted that for solutions of LiClO4 in acetone-nitromethane mixtures, the 935-cm-1 Raman band (vl, symmetric stretch) of the perchlorate ion ramained narrow and at constant frequency in solutions with acetone/Li+ mole ratios of :4. As the acetone/Li+ mole ratio became less than four, the band broadened and shifted to higher frequency. From these data and from the 35Cl nuclear magnetic resonance results that were 54 discussed earlier in this chapter, it is evident that for lithium perchlorate-acetone solutions in nitromethane in which the acetone/Li+ mole ratio is 1:1, the predominant solute species is the Li(acetone)+ClOZ ion pair. The pur- pose of this investigation was to add a weak complexing agent to solutions in which the acetone/Li+ mole ratio was 1:1 and then observe if it displaced the perchlorate ion from the solvation shell of the lithium ion. For this purpose, pentamethylenetetrazole (PMT) was used. The structure of PMT is shown below. N=C—CH2-——CH2 Z / N—N—CHZ— CH2 Previous studies (70) by 7Li NMR have shown that in nitro— methane solutions lithium forms a 1:1 complex with PMT l (pentamethylenetetra— with a formation constant of 4.85 M— zole will be discussed in detail in the second part of this thesis). Far-infrared spectroscopy was used to confirm the formation of this complex. For a nitromethane solu- tion that was 0.5 M in LiClO4 and 1.5 M in PMT, a broad band was observed at 408 cm—1 that could not be attributed to the salt, solvent, or complexing agent. Thus, this . . . .+ band was aSSigned to be due to a Vibration of the Li - PMT complex. 55 To investigate the effect of PMT on the Li+ClO4 ion pair formation, the 935-cm-l Raman band of perchlorate ion was monitored for solutions in which the acetone/Li+ mole ratio was 1:1 as increasing amounts of pentamethylenetetra- zole were added. From the results shown in Figure 11, it is seen that as the concentration of PMT is progressively increased, the band changes from a broad and fairly weak peak to a very sharp, intense peak which is indicative of the free perchlorate ion. Therefore, pentamethylene- tetrazole displaces the C104 ion from the solvation shell of the lithium ion according to the equilibrium Li(Acetone)+ClO4 + PMT Z Li(Acetone)(PMT)+ + C104 (5) At the same time only a very weak acetone band was 1 observed at 789 cm— , while a larger band was located at l 803 cm- , which indicated that most of the acetone was in the complexed form. Therefore, although PMT displaces C10; from the Li+ solvation shell, it does not displace any acetone from the solvation shell of the lithium ion. 56 3:1 2:1 1:1 1 1 1 1 l 960 950 940 930 920 cm“ -. m *1 . — Figure 11. rho 935~cm C104 Raman band as a function of the PMT/LiClO4 mole ratio. PART II GAS CHROMATOGRAPHIC STUDIES OF CONVULSANT TETRAZOLES 57 CHAPTER I HISTORICAL 58 The chemistry of tetrazoles and substituted tetrazoles has been the subject of investigations for many years be— cause of their strong stimulating action on the central nervous system. The parent compound, tetrazole, is composed of one carbon and four nitrogen atoms in a ring and can exist in two tautomeric forms I and II (71,72). H H H \Nl—-C/5 lN-_—_-c5 2 \\ 4 2 / 4 N N H—N N \N / \N/ 3 3 I II It has been shown that about 97% of the equilibrium mixture of I and II exists in form I (73). The numbering of the atoms in the ring starts with the nitrogen atom adjacent to the carbon (configuration I) and proceeds counterclock- wise around the ring as shown above. The hydrogen atoms in positions 1 and 5 can be easily replaced to give rise to two classes of 1,5-disubstituted tetrazoles. In the first of these, both hydrogens are replaced by aliphatic or aromatic groups as shown below. ’20 R 1 2 / ——c 2 \Z / \\N / N 1,5—disubstituted tetrazole 59 A special class of 1,5-disubstituted tetrazoles is composed of the cyclopolymethylenetetrazoles in which a hydrocarbon chain forms a second ring that is fused to the tetrazole ring. / \\5 n=3,4,5,... N 4 cyclopolymethylenetetrazole The cyclopolymethylenetetrazoles and their derivatives are the primary subject of this investigation. The cyclopolymethylenetetrazoles and their deriva— tives are known for their ability to cause epileptic con- vulsions. The convulsant activity varies with the number of methylenes in the hydrocarbon chain as well as with the nature and position of substituent groups (74). The con- vulsant activity and some physical properties of various tetrazoles are listed in Table 6. As can be seen from Table 6, the insertion of methylene groups into the hydrocarbon ring significantly changes the convulsant property of the drug. A gradual monotonic increase in convulsant activity results as the number of methylene groups is increased. Thus, this series of com- pounds seems very useful for the study of possible _. "' . 61 .HmEHCM on» NO panoB Soon mo EmuooHHx Hod oHoN Imuuou mo mEmHmHHHNE mo mpHco CH co>Hm wH mommOU one .oHSNme mo mEODQESm umHHw ecu omsmo 0p Summmoooo mHonHuop mo ucooEm EsfiflcHE onu mH mmMmOp ucmmH5>QOU EDEHCHE ore k. om ON mm.va 0HonHumuocoawcuofimucomHSusnlmmmum m mmH ON.¢OH 6HonsomesmHNsmemosmmHNosououmuIO oHQsHOmcH mm vm.mmm oHonnuouoclonuoEwoopco wHQsHOmcH om mm.va mHoNoHuouocoawguoEmcoz oHQSHOon NHH om.omH oHonuuouoconnqumuoo om NV mm.mmH mHonyuouoconnuoEoumwm O4 NO ON.NmH wHonuomowcwHNsnwssxwm om OO OH.NNH mHoNssomomsmHNsowsmosms omN NHH mH.sNH wHonuomowsmHNsomsmsome OOOH OHH NH.OHH mHonsnmomcwHNsumsHue Remy H.omOnH Hoov ucfiom Dcmfloz oHonHuoB pcme5>ooo ESEHQHZ mcHuHoz Hmasooaoz .moHoNouuouocoawzquwHomoaomu oEom mo moHuHomoum HMUHmoHoomEHosm pom HMQmezm .m OHQMB 62 correlations between physicochemical properties and phar- macological activity. Of the cyclopolymethylenetetrazoles, only penta- methylenetetrazole (PMT) has had some clinical applications. It has been used in patent medicines as a respiratory and cardiac stimulant, and in higher doses it has been used as an analeptic in barbiturate overdoses. Pentamethyl- enetetrazole has also been employed for screening anti— convulsant drugs and in veterinary medicine in hastening the recovery of animals from anesthesia. Due to its use in chemotherapy, the chemistry of PMT has been studied in some detail. However, the analytical chemistry of PMT and the other cyclopolymethylenetetrazoles has not been thoroughly investigated. A large number of studies have been devoted to the investigation of several PMT—transition metal complexes. These complexes have been discussed by Dister (75) and Popov (76), and as a result of these studies, considerable physical data pertaining to these compounds have been ac- cumulated. Complexes such as CuCl'PMT (77), HgCl °PMT (78), and 2 CdC12:PMT (79) have been widely investigated. Because of their moderately low dissociation constants and solubili- ties, these complexes were employed in precipitation pro- cedures for the quantitative analysis of pentamethylene- tetrazole. There are some disadvantages, however, in using these complexes for the gravimetric determination 63 of PMT. The conditions of the reaction must be rigorously controlled so that the exact composition of the precipi- tate is known, as the ratio of PMT to metal can vary from one to two. Another drawback is that the solubility of the complexes seriously limits the analysis at low con— centration levels. Nonaqueous potentiometric determinations of PMT and substituted pentamethylenetetrazoles in formic acid were reported by Popov and Marshall (80,81). They determined the tetrazoles quantitatively and noted that the tetrazole with the greatest convulsant activity was the most basic and the most inactive compound was the least basic. How- ever, the elaborate procedure used does not make this method practical for routine analytical determinations. Beyrich and Schlaak (82) titrated PMT with perchloric acid-glacial acetic acid mixtures in a vessel containing benzene and acetic anhydride. They determined penta- methylenetetrazole in the presence of other drugs at con- centration levels of 0.3 to 3.0 parts per thousand. A spectrophotometric method for the determination of PMT in pharmaceutical preparations was described by Daoust (83). The analysis was based on the precipitation of the CuCl'PMT complex which was isolated and dissolved in nitric acid. The copper was then complexed with tetraethylene- pentamine and the absorbance of the solution measured. The relative standard deviation was reported to be 8% at the 40 parts per million level. Turczan and Goldwitz (84) measured the concentration of PMT in pharmaceutical preparations by proton nuclear magnetic resonance spectroscopy. Their reported detec- tion limits were very poor as they found that the best results were obtained with solutions containing at least 3% by weight PMT. Rylance and co-workers (85) investigated the use of thin layer chromatography in the determination of neutral drugs and found that PMT could not be determined due to its lack of an appreciable ultraviolet absorption near 254 nm. However, Guven (86) found that PMT could be spotted with a mixture of 10% copper sulfate and 2% ammonia solu— tions which gave blue spots upon drying. Gas chromatography was first employed in the determina- tion of pentamethylenetetrazole by Kawamoto (87) in 1962. His instrument was equipped with a thermal conductivity detector, and the baseline drift obtained for aqueous solu- tions was so severe that the determination of the peak areas was extremely difficult. Several investigators have recently reported the use of gas chromatography with flame ionization detection systems for the analysis of PMT (88—90). Generally, liquid stationary phase coatings of 3—5% have been used. Penta- methylenetetrazole has been routinely determined in the 100 parts per million level with the technique capable of measur— ing concentrations as low as 10 ppm. Clearly, this is the best technique thus far reported 65 for the quantitative determination of PMT since the analysis time is fast and the sensitivity is excellent. Although this method has been successfully applied to the analysis of PMT, there still are no reports on the analysis of the other cyclopolymethylenetetrazoles. Hence a portion of this research is involved with the quantitative determina- tion of the cyclopolymethylenetetrazoles by using gas chromatography. Numerous attempts to correlate physicochemical prop- erties with physiological activities of the tetrazoles and other biologically active compounds have been made. In addition, several studies have centered around the deter- mination of the nature of the tetrazole interaction in the biological system. Popov and Holm (91) determined the dipole moments of pentamethylenetetrazole, B-EgEEfbutylpentamethylenetetra- zole, and 8-gggfbutylpentamethylenetetrazole in benzene solution to be 6.14, 6.20, and 6.18 D, respectively. They concluded that there was no correlation between their convulsant activities and the magnitude of their dipole moments. Schueler gg‘gl. (92) studied correlations of the bio- logical activity of some substituted tetrazoles with their ultraviolet absorptions. They found that alkyl-substituted tetrazoles of moderate activity generally showed little or no absorption down to 220 nm. However, aryl-substituted tetrazoles which act as depressants showed absorption bands 66 in the 290 and 225 nm spectral regions. Apparently, there is some correlation between ultraviolet absorptions of the tetrazoles and their physiological activities. Erlich and Popov (93) determined basicity constants for six cyclopolymethylenetetrazoles varying from trimethyl— enetetrazole to undecamethylenetetrazole in formic acid solutions. It was shown that while the cyclopolymethylene- tetrazoles do not have any detectable proton affinity in aqueous solutions, they do act as fairly strong monoprotic bases in formic acid solutions. The authors did not ob- serve any correlation between the length of the hydrocarbon chain and the base strength of the tetrazole ring, as nearly all of the reported pr values were about 1.8. The surface activity of PMT was investigated by Buchanan gg gl. (94) in a study of air—solution surface tension isotherms. They observed that central nervous system stimulants prefer the aqueous bulk phase, while drugs which exhibit depressant action collect at the air— solution interface. Recently, it has been shown that PMT can emulsify human cell membranes (95). This action weakens the membrane and causes it to rupture. Due to its solubility in lipid substances, PMT then diffuses rapidly through the membrane. Gross and Woodbury (96) studied the effects of various cyclopolymethylenetetrazoles on ion transport in toad bladder membranes. They noted a strong correlation between the convulsant potency of the tetrazoles and the increase -——-— -u—u-- .. L 67 of the short-circuit current produced in the isolated toad bladder. It was concluded that the cyclopolymethylenetetra- zoles affect the potassium ion transport across the membrane. One possible explanation for this action is that the tetra- zoles have an effect within the membrane. If this explana- tion is correct, the tetrazole must first pass into or through the membrane. One theory for the passage of materials through a membrane postulates an actual dissolution of the material in the membrane. Thus an investigation of this phenomenon must involve studying the partitioning of the substance between an aqueous solution and the lipid membrane. There have been no reports, however, on the partition coefficients of the cyclopolymethylenetetrazoles, although investigations have been carried out in which physicochemi- cal properties have been related to the partitioning of other biologically active compounds between aqueous solutions and lipid solvents. Meyer (97) and Overton (98) showed that the relative narcotic activities of drugs often paralleled their oil/ water partition coefficients. They also noted that in a homologous series of compounds the partition coefficient increased by a factor of from two to four per methylene group. Recently, partition coefficients have been used as extrathermodynamic reference parameters for "hydrophobic bonding" in biochemical and pharmacological systems (99, 100). 68 Although partition coefficients have been tabulated for many biologically important compounds (101), very little is known about the distribution of the cyclopoly- methylenetetrazoles between aqueous solutions and lipid solvents. Thus another section of this thesis is devoted to the investigation of the partition coefficients of the cyclopolymethylenetetrazoles. CHAPTER II EXPERIMENTAL 69 REAGENTS Carbon tetrachloride (Fisher) was shaken with alcoholic sodium hydroxide and washed several times with water. It was then dried over calcium chloride and fractionally dis- tilled. Water was doubly distilled in an all glass apparatus, once with potassium permanganate present in the charge to remove all oxidizable organic impurities. Chloroform (Fisher, Certified A.C.S.) was used without further puri- fication. Pentamethylenetetrazole (Aldrich) was recrystallized from diethyl ether and dried £2 ggggg. Trimethylenetetra- zole (Aldrich) was purified by recrystallizing about 10 grams of the tetrazole from a solvent mixture of 50 ml of carbon tetrachloride and 10 ml of ethanol. The other cyclopolymethylenetetrazoles were prepared and purified as described by D‘Itri (102,103). Analytical reagent grade sodium chloride (Mallinckrodt) was used without further purification. SYNTHESIS OF 8-SEC-BUTYLPENTAMETHYLENETETRAZOLE 4-sec-butylcyclohexanol To prepare 8-sec-butylpentamethylenetetrazole, 4-sec- butylcyclohexanone was needed as the starting material. This compound, however, is not commercially available, and the most similar compound that is commercially 70 71 available is p—ggg-butylphenol. This phenol was used as the starting material for this synthesis. The hydrogenation reaction was carried out in the Parr Series 3910 Hydrogenation Apparatus which can withstand pressures of up to five atmospheres. The experimental procedure was similar to that reported by Somerville and Theimer (104) for the hydrogenation of p-gggg-butylphenol. One hundred grams of p-ggg-butylphenol (Eastman Organic Chemicals, Practical Grade) were dissolved in 100 ml of absolute ethanol in the reaction bottle, and 4 grams of 5% rhodium on alumina catalyst (Pfaltz & Bauer) were added. The reaction bottle was mounted in the Parr apparatus and subjected to a pressure of 50 psi of hydrogen. The reaction vessel was shaken and heated to about 80°C. As the pressure in the reservoir decreased, more hydrogen was added to maintain a pressure of 50 psi. The reaction was allowed to proceed for several days until there was no further uptake of hydrogen. The apparatus was then disassembled and the reaction mixture was fractionally distilled at reduced pressure. The product was collected at 120-130° at 20 mm pressure [lit. (105) 128° at 20 mm]. The yield of 4-ggg—butylcyclo- hexanol was about 50%. 4—sec—butylcyclohexanone The oxidation of the 4—sec-butylcyclohexanol was Carried out by using the procedure for the preparation of 72 menthone from menthol (106). A solution of 30 ml of concentrated sulfuric acid in 340 ml of water was added to 68 grams of sodium dichromate in a one liter round—bottomed flask. Fifty grams of 4-ggg— butylcyclohexanol were then added in three portions while the mixture was stirred. Heat was evolved and the tempera- ture of the reaction mixture increased to approximately 55°. After the mixture cooled to room temperature, the oil was mixed with an equal volume of ether, separated in a separatory funnel, and washed with three lOO—ml por- tions of 5% sodium hydroxide solution. The ether was re- moved and the residue distilled under reduced pressure. The product was a colorless liquid which distilled at 110-120° at 30 mm [lit. (105) 104-106° at 13 mm]. Twenty- five grams of the 4-sec-butylcyclohexanone were obtained. 4-sec-bugylcyclohexanone Oxime The cyclohexanone was converted into the oxime by treatment with an aqueous solution of hydroxylamine as described by Herbst and co-workers (107). A mixture of 15.4 grams of 4-ggg-butylcyclohexanone and 8.4 grams of hydroxylamine hydrochloride (Matheson, Coleman, and Bell) was added to 70 ml of a 10% sodium carbonate solution. The mixture was stirred for a few hours and the oxime was extracted with ether. The ether was removed and the product was distilled at 140-1450 at 28 mm. The yield of the product, a colorless liquid, was 73 approximately 85%. 8-sec—butylpentamethylenetetrazo1e The 8—ggg—butylpentamethylenetetrazole was prepared by using the method described by Herbst and co-workers (107) for the preparation of 8-isopropylpentamethylene- tetrazole. A suspension of 14.6 grams of powdered sodium azide (Eastman Organic Chemicals, Practical Grade) in 250 ml of 1,2—dichloroethane was added to a 2—liter three-necked flask equipped with a stirrer, dropping funnel with the tip im— mersed in the reaction mixture, exit tube, and a long—stemmed alcohol thermometer with the bulb immersed in the reaction mixture. CAUTION: The following procedure involves the genera- tion of hydrazoic acid. Hydrazoic acid vapors are highly toxic, and all reactions in which it is involved should be carried out in an efficient hood. Heavy metals, such as mercury from a broken thermometer,must be excluded because of the explosive nature of mercury (II) azide which could form. While stirring the suspension, 120 grams of chloro- sulfonic acid were added at such a rate that the tempera— ture did not rise above 35°. After the addition of the acid was completed, a solution of 16 grams of 4-ggg-buty1- cyclohexanone oxime in 125 ml of 1,2-dichloroethane was added with continuous, vigorous stirring so that the 74 temperature remained between 35 and 45°. Upon complete addition of the oxime, stirring was continued until the reaction mixture had cooled to room temperature. Water was then slowly added to the mixture to decompose the excess chlorosulfonic acid (external cooling was required). The aqueous layer was separated and the acid neutralized with aqueous sodium hydroxide. The neutral solution was ex— tracted with four lOO—ml portions of 1,2-dichloroethane. The extracts were then combined with the 1,2—dichloroethane solution, dried over sodium sulfate, and the solvent removed by evaporation. The residue was boiled with 100 ml of 10% aqueous hydrochloric acid for three hours, and the product was extracted with several portions of 1,2-dichloroethane. After washing the combined extracts with water and drying over sodium sulfate, the solvent was removed by evaporation. The product, a dark brown oil, was then cooled for several hours in an ice bath in order to facili- tate crystallization. The crude product was purified by recrystallizing it a few times from a mixture of heptane and 1,2-dichloropropane. The melting point of the purified 8-ggg-butylpentamethylenetetrazole was 69-70° [lit. (107) 70-7l°]. The yield based on 4—ggg-butylcyclohexanone oxime was approximately 50%. The steps in the synthesis of this compound are summarized in Figure 12. 75 H2 gm \_ Rh on Alumina OH p Cr207 @ NHon < NOH o CHg-CIZH-CHfi-CH3 C HZT/ {HuHmcmm o>HpmHom o>HHMHom .moHoN ImuuouoooamzpofimaomoHomu ocu How moHuH>HuHmcom paw moEHB soHpcopom 0>Hpmamm .N oHooe 86 II .mmHonuuwumcmHanuoESHomoHomo onu mo uanoB HmHsooHoE on» m> oEHu coflucouou 0>HumHoH may no uon .mH ousmHm F1653 m<430w402 com com on. 00. ,1 _ cod no He 00. w A 3 ..oo.~ m .I‘ 3 . w. loom m N n... 109. m 87 the hydrocarbon chain increases. This same observation has been made for members of other homologous series such as paraffins, ketones, and esters (110). In this respect, therefore, the behavior of the cyclopolymethylenetetra- zoles is that which is expected of any regular homologous series of compounds. The chromatographic sensitivities of the cyclopoly- methylenetetrazoles were measured for single component solutions in chloroform. These values, listed in Table 7, were determined by dividing the average detector response by the number of nanomoles of sample injected. Since changes in the instrumental parameters can greatly affect the detector response, the sensitivities are also reported relative to that of PMT. However, the relative detector response can be measured either in terms of peak height or peak area; but since the concentration of the sample is proportional to the peak area, one should probably give more significance to the values determined on this basis. Upon examining the relative sensitivities based on peak area in Table 7, a near linear increase is observed as methylene groups are added to the hydrocarbon chain. The glaring exception is trimethylenetetrazole which has a very low value of 0.16. This compound was investigated in more detail, and it was determined that trimethylene- tetrazole was undergoing thermal decomposition. By lower- ing the temperature of the injection port to 175°, this 88 decomposition was reduced, but was not completely elimin- ated. RESOLUTION OF A MIXTURE OF TETRAZOLES A solution containing a mixture of the cyclopoly- methylenetetrazoles, each present at a concentration of 0.0138 grams per 10 ml of chloroform, was prepared to determine how well the individual components could be resolved. As can be seen in Figure l4,base1ine resolution is achieved in most cases, and the only problem encountered is that trimethylenetetrazole, tetramethylenetetrazole, and pentamethylenetetrazole are all eluted at the same time. This behavior was not unexpected since the relative retention times of these compounds are very similar. Thus, to separate these three tetrazoles, one must either use a different column packing or reduce the temperature of the column oven. The other cyclopolymethylenetetrazoles are quite well resolved, however, and the relative retention times mea- sured for the components in the mixture are identical to those measured for the single component solutions previously discussed. Thus, the retention times are not influenced by the presence of other species in solution and can be used to identify qualitatively these cyclopolymethylene- tetrazoles. Figure 14. Chromatogram of a mixture of cyclopolymethyl— enetetrazoles. Sample: 2.0 ul of C3 - Cll tetrazoles, each present at a concentration of 0.0138 g/lO ml of chloroform. Injector: 220°; Column: 210°; Detector: 300°; He at 60 ml/min; Sensitivity: 128 x 10"11 a.f.s. 90 Figure 14 C3C4C5 Co C0 C9 * Cll ngurrgv: Jo 1'2 ‘l‘ '2 MINUTES 91 WORKING CURVES To determine the cyclopolymethylenetetrazoles quanti- tatively, it was necessary to construct working curves for each of these compounds. It was observed that for tetrazole solutions up to a concentration of 0.1 M, the peak shape was symmetric and reproducible, with no broaden- ing at the higher concentrations. Thus, the calibration curves were prepared by plotting the average detector re— sponse (in units of peak height) versus the concentration of the tetrazole. A typical working curve is illustrated in Figure 15. In this case the average detector response was monitored as a function of the PMT concentration. The PMT concen- tration ranged from millimolar to 0.1 M, and as can be observed in the figure, the plot is linear over this concentration range. Similar working curves have been obtained for the other cyclopolymethylenetetrazoles in this concentration range. At concentrations greater than 0.1 M, however, the curves begin to deviate from linearity. This behavior is illustrated in Figure 16, where a working curve for heptamethylenetetrazole is shown. In this study the con- centration ranged from 0.02 to 0.5 M. As can be seen in the figure,at the higher concentrations, the curve bends towards the concentration axis. Despite this deviation from linearity, these calibration plots are still useful 92 Figure 15. Plot of average detector response XE concentra- tion of pentamethylenetetrazole in aqueous solution. 93 l C) 0) l l l .1 C3 C) N v— 891 x asNOdsau sow-31w Figure 15 L 0.06 0.08 0.10 CONCENTRATION OF PMT (M) 1 0.04 1 0.02 94 .coHumHom oUHHOHnomuuou cooumo CH oHonuuouoconSHofimumon mo cofluouucoocoo m> omcommmu Houomump oomuo>¢ as 50.0 to 20.25.2328 mO HO NO NO 8 .OH wusmHm _ _ _ — l CD cu Ol X HSNOdSEU HDVHSAV 0v. 95 for the analysis of tetrazoles at these higher concentra- tions. As a result of this investigation, a gas chromatographic method for the routine analysis of the cyclopolymethylene- tetrazoles at the 50-100 ppm level has been developed, and the technique is capable of determining concentrations of less than 10 ppm. CHAPTER IV SOLUBILITY AND DISTRIBUTION STUDIES OF THE CYCLOPOLYMETHYLENETETRAZOLES 96 SOLUBILITY STUDIES The solubilities of several cyclopolymethylenetetra— zoles were determined in aqueous solutions, and the data are presented in Table 8. As expected, an increase in the number of carbon atoms in the polymethylene chain generally decreases the solubility in aqueous solution. The glaring exception, however, is pentamethylenetetrazole which is soluble in water up to a concentration of 5.0 M. Table 8. Solubilities of some Cyc10polymethylenetetrazoles Molal Solubility Tetrazole Solubility in g/ml Trimethylenetetrazole 1.4 T 0.16 Pentamethylenetetrazole 5.0 0.69 Heptamethylenetetrazole 0.18 0.031 8-ggg-butylpentamethylenetetrazole 0.0052 0.0010 B-EgEE-butylpentamethylenetetrazole 0.0029 0.00057 It has been reported previously that PMT is essentially completely miscible with water, as viscous solutions with concentrations up to 750 grams of PMT per 100 grams of water were obtained (111). The present study indicates that aqueous solutions of PMT have a strong tendency to supersaturate. 97 98 It is also interesting to note the solubilities of the two isomeric pentamethylenetetrazole derivatives, namely 8-gggfbutylpentamethylenetetrazole (B-gggfbutyl PMT) and 8-gggg-butylpentamethylenetetrazole (8‘EEEE' butyl PMT). While the structural differences between the two compounds are very minor, the water solubility of 8- sec-butyl PMT is nearly twice that of the other isomer. DISTRIBUTION STUDIES Since there is a possibility that the physiological action exhibited by the cyclopolymethylenetetrazoles may be due to an interaction of the tetrazole with a lipid membrane, it was decided to study the distribution of these tetrazoles between aqueous solutions and lipid sol— vents. In addition to investigating the series of cyclo- polymethylenetetrazoles, it was also of interest to study some pentamethylenetetrazole derivatives. The compounds 8-ggg—butyl PMT and 8-gggg-butyl PMT were of particular interest because the convulsant activities of these two tetrazoles were reported by Gross and Featherstone (112) to be 750 milligrams per kilogram of body weight for 8- .gggfbutyl PMT and only 3 milligrams per kilogram of body weight for B—Eggg-butyl PMT. Thus a small change in the nature of the substituent group greatly affects the phar- macological activity of these compounds. However, it was 99 recently shown that the convulsant activity of 8-ggg— butyl PMT is 50 milligrams per kilogram of body weight rather than the 750 milligrams that was reported previously (Appendix II). An appropriate solvent for these distribution studies should have the solvent properties of lipids and have a very low water content on saturation. Carbon tetrachloride fits these requirements. In addition it is readily avail- able and can be purified easily. Thus carbon tetrachloride was selected as the nonaqueous solvent for these studies. The distribution ratios of pentamethylenetetrazole, hexamethylenetetrazole (C6CMT), and heptamethylenetetrazole (C7CMT) between water and carbon tetrachloride solutions were determined at several concentrations. The distribu- tion ratio, D, is defined as the total concentration of the tetrazole in the organic phase divided by the total concentration of the tetrazole in the aqueous phase. The results are listed in Table 9. In all three cases the distribution ratio is somewhat concentration dependent. Thus for all three of these com- pounds, this concentration dependence may be an indication that some reaction such as dimerization is occurring in one or both of the phases. As can be seen in Figure 17, the distribution ratio for PMT increases gradually from 0.18 to 0.33 over the concentration range of 0.1 to 0.5 M. Linear plots of the distribution ratio versus the concen- tration of the tetrazole are obtained for C6CMT and C7CMT 100 m.oamv.m mmvo.o mmv.o m.o H.OHNo.m omvo.o mwm.o v.0 m.OHNO.N ONm0.0 OON.O m.O N.oHNv.m ommo.o mNH.o N.o m.ono¢.m mmmo.o wmbo.o H.O mHONmuuopmcoamnpoEmumom mO.OHOH.N HOH.O mem.O m.O wo.onN.H ovH.o omm.o H.O mO.OHN¢.H mNH.O mNH.O m.O vo.OHmo.H mmo.o OOH.o m.o O0.0HNN.O Nm0.0 OHO0.0 H.O mHonuuoumcmHNsomEMxmm HO.OHmm.O ONm.O mNH.O m.O NO.OHON.O OON.O mmO0.0 H.O HO.OHmN.O ONN.O OOO0.0 m.O Ho.owmm.o mmH.o memo.o I N.o HO.OHNH.O ONO0.0 OOH0.0 2 H.O wHonuomumcmHNsomsmosme oflpmm ommnm mmonm coauomuuxm oHonuuoB soHuanuumHo on sH sHoo sH macemm . OGOU . OGOU . OGOU .wcoHumHucooaoo mooHHo> pm oHONmuuouocmawnpoEmumom poo .oHonHpouoconnuofiwxmm .oHonHumuocoHSonEmusom mo moHumm coauonwuumHo .m oHQme 101 Figure 17. Plot of the distribution ratio 2g the concentra- tion of pentamethylenetetrazole for the water- carbon tetrachloride system. 102 00 $13 548425525525qusz “.0 29.552828 to 00 NO 9 _.O 0.0 u H _ 1.) 0.0 1_.0 1 NO 1 md 1 v.0 0 Figure 17 103 (Figures 18 and 19). Distribution studies were also carried out on some other cyclopolymethylenetetrazoles. In the case of 8- Egggrbutyl PMT, the compound was only soluble to a concen- tration of 0.05 M,in carbon tetrachloride and only slightly soluble in water. Due to this limited solubility it was very difficult to determine the concentrations of the tetra— zole in the aqueous and organic phases. In this case the distribution ratios were determined from the chromatograms as the average detector response of the organic phase divided by the average detector response of the aqueous phase. The results are listed below in Table 10, and a plot of the distribution ratio versus concentration is shown in Figure 20. Once again, the distribution ratio is dependent on the concentration of the tetrazole. Table 10. Distribution Ratios for 8-tert-butylpentamethyl- enetetrazole between Water and Carbon Tetrachloride Concentration Before Extraction Distribution Ratio 0.005 M 26.1 0.02 28.1 0.04 31.9 0.05 35.6 104 .Eoummw oUHHOHMomuuou cooumouumum3 map How oHoNouuoumconnuoonom mo COHumuucmocoo m> oHumu :oHuooHuumHo .OH wusOHs 33 5522552355515... .6 2255250200 mHV .VHV mHu «Ho HHV H H H _ ogu 105 .Emuwhm oUHHOHmvauou cooumounoHMB map How oHonuuouoconzuoEmumo: mo coHumuucoocoo m> oprH coHuanHumHa :13 3022552355255... .5 2225250260 no Yo no Nd H.O .OH musmHm 0.0 H H H H 0.0 u ON .1 0.2 106 .Emummm opHuoHnomuuou cooumosuouws on» How oHoN ImuuouosoaanuoeoucomHmusoluuounm mo coHumuucoocoo MM oHuwH coHusoHHumHa .om ousmHm 3.5 :2“. 355-226 .10 292525028 I l 00.0 V0.0 ”0.0 N0.0 H0.0 00.0 H d H H HOGN 0.0» 0.0» 107 Attempts were also made to determine the distribution ratios for trimethylenetetrazole (C3CMT) and 8-gggfbutyl PMT. In the case of C3CMT, the distribution ratio appeared to be extremely small, which, when coupled with the very low chromatographic sensitivity of trimethylenetetrazole, made it very difficult to analyze the organic phase. How- ever, from the studies performed it was estimated that the distribution ratio for C3CMT is “0.0014 at a concentration of 0.20 M. On the other hand, the distribution ratio for 8-ggg- butyl PMT was so large that it was difficult to analyze the aqueous phase. It was estimated that the distribution ratio for 8-sec-butyl PMT is on the order of 220 at a concentration of 0.20 M. DISCUSSION OF RESULTS A summary of the distribution ratios of the various cyclopolymethylenetetrazo1es for the water-carbon tetra— chloride system is shown in Table 11. In comparing the values for trimethylenetetrazole through heptamethylene- tetrazole, it is observed that the distribution ratio in- creases as methylene groups are added to the hydrocarbon chain. This relationship was not unexpected since it was previously reported that in a homologous series of compounds, the partition coefficient usually increases by a factor of two to four for each additional methylene group (97,98). It is also interesting to note that at least for PMT, 108 Table 11. Distribution Ratios of some Cyclopolymethylene- tetrazoles between Water and Carbon Tetrachloride. Conc. Before Distribution Tetrazole Extraction Ratio Trimethylenetetrazole 0.20 M 0.0014 Pentamethylenetetrazole 0.20 0.22 Hexamethylenetetrazole 0.20 1.05 Heptamethylenetetrazole 0.20 5.47 8-tert—butylpentamethylenetetrazole 0.05 35.6 8-sec-butylpentamethylenetetrazole 0.20 220 109 C6CMT, and C7CMT, the distribution ratio depends on the concentration of the tetrazole. Thus for all three com- pounds, it appears that some kind of association reaction may be occurring. In order to describe quantitatively the partitioning of these compounds, it was of interest to determine the nature of this reaction. The possibility of an acid-base reaction in the aqueous phase can be excluded since the addition of cyclo- polymethylenetetrazoles to an aqueous solution does not result in any change of the pH of the solution (93). Therefore, the only possible reaction involves the forma- tion of dimers or higher aggregates in one or both of the liquid phases. To simplify the calculations it was first assumed that the variation of the distribution ratio with concen- tration was due to the formation of dimers in the organic phase. It was also assumed that only the tetrazole monomer was being transferred from one phase to the other. The dimerization reaction can be represented by the equilibrium 2ummno p0>uomoo .ocou .mpcosouommoz 0Hmoomo>nu an pocHEHmuoo mm COHuoHom mooosom :H moHonpoouoconnpoE>HomoHomo 080m mo mpanoz HMHsooHoz .NH oHome 115 likely a very complicated system which involves several complex equilibria, however the calculated partition co- efficient of 0.12 is probably a reasonable estimate of the actual partition coefficient. For hexamethylenetetrazole and heptamethylenetetrazole, however, the cryoscopic data do not indicate that any assoc- iation is occurring in the aqueous phase. Thus for these compounds, the assumption that no association is occurring in the aqueous phase appears to be valid; but as noted previously, however, the distribution data for C CMT and 6 C7CMT do not fit the derived equation that involved the formation of dimers in the organic phase (Equation (10)). It was then assumed that these compounds formed primarily trimers in the organic phase. The trimerization reaction is given by 3(Tz)org I (Tz)§rg (11) with the corresponding equilibrium constant Tz org _ ( )3 (12) trimer. 3 (T2) From these equations and from Equations (8) and (9), the following relationship was derived. D = 3 Korg p3 (Tz)2 + p (13) trimer. aq 116 If Equation (13) is valid, a plot of the distribution ratio versus the square of the concentration of the tetra— zole in the aqueous phase should produce a straight line. The intercept should be equal to the partition coefficient, org trimer could be calculated from the P, and the value for K slope. For both C6CMT and C7CMT, plots of D versus (T2):q yield very good straight lines as can be observed in Fig- ures 22 and 23, respectively. In the case of C6CMT the org partition coeffic1ent was determined to be 0.5 and Ktrimer. was calculated to be 170; while for C7CMT the partition co- efficient was found to be 1.6 and the trimerization equilib- rium constant was determined to be 310. For these two tetrazoles, since the distribution ratio increases linearly with increasing tetrazole concentrations (Figures 19 and 20), there is another method by which the partition coefficient could be estimated. This method involves the extrapolation of the line back to zero con- centration. At this point the distribution ratio and the partition coefficient, P, should be equal. In this manner, the partition coefficients were (determined to be 0.34 for C6CMT and 2.1 for C7CMT. These values compare quite favorably to those obtained by using Equation (13), as can be seen in Table 13. 117 Figure 22. Plot of the distribution ratio 1g the square of the concentration of hexamethylenetetrazole in the aqueous phase for the water-carbon tetrachloride system. 118 QNOAV 306 H 68.6 d on O .30 OH owOHu NPOHV DOOHu QOOAV OHH (H [OH I o.w Figure 22 119 .Emumwm opwuoanomnuou cooumouuoum3 030 How woman msoooUMIocu CH oHonuuouogo (Hmnuwewumoz mo coHpmHucoosoo may no mumsvm may m> oHumH soHuooHuumHo .mm onsmHm .MTE oNoH 6866 flood «Sod ooood eoood o.o . H H _ H . _ oo 0.9 120 Table 13. Partition Coefficients for Hexamethylenetetra- zole and Heptamethylenetetrazole. P Obtained P Obtained Tetrazole by Extrapolation by Equation (13) Hexamethylenetetrazole 0.34 0.5 Heptamethylenetetrazole 2.1 1.6 Since these results are in fairly good agreement, it would seem that the values of the partition coefficients for C6CMT and C7CMT between water and carbon tetrachloride obtained in this investigation are, indeed, good estimates of the actual partition coefficients. However, whether or not these compounds actually form trimers in the organic phase is still not entirely clear, as no supporting data have been obtained. Moreover, a mechanism has yet to be postulated for the formation of trimers. Thus additional investigations are in order so that the nature of the solute species in carbon tetra- chloride solutions may be determined. Although the results obtained in this investigation did not lead to any definite correlations with the phar- macological prOperties of the cyclopolymethylenetetrazoles, it would be interesting to extend these distribution studies to include other solvent systems to determine if any such correlations exist. APPENDICES APPENDIX I DESCRIPTION OF PROGRAM RAMAN AND THE ACCOMPANYING INTERFACE FOR THE SPEX RAMALOG 4 RAMAN SPECTROMETER 121 DESCRIPTION OF PROGRAM RAMAN AND THE ACCOMPANYING INTERFACE FOR THE SPEX RAMALOG 4 RAMAN SPECTROMETER RAMAN is a program written in FORTRAN and SABRE de- signed to acquire data from the Spex Ramalog 4 Laser-Raman spectrometer equipped with the appropriate interface. The program was written and the interface designed by Patrick M. Kelly of this department. The program features include: 1. acquisition of up to 500 data points at 0.1 or 1.0 wavenumber intervals; 2. signal averaging of each data point; 3. correction for baseline drift; 4. an option of integrating any spectral bands; 5. presentation of data in tabular form. INTERFACE Triggering Interface The Spex Ramalog 4 spectrometer has a wavenumber en— coder which makes a 2.5 v pulse available at a rate of one pulse per tenth wavenumber. The frequency of the pulses during a scan is further divided by using a series of decade counters so a pulse may be obtained every ten 122 123 or one hundred wavenumbers. This interface taps the ap- propriate decade counter in order that a pulse can be obtained either at a rate of one per wavenumber or one per tenth wavenumber. These pulses are then used to trigger a Schmidt trigger which initiates A/D conversion at the appropriate time. The Schmidt trigger is then reset under software control. The two encoder signals are used to trigger data acquisition in the following manner: Mode 1: Data acquisition at 0.1 cm.1 intervals Since the spectrometer is usually scanned at a rate of 20 wavenumbers per minute, encoder pulses will never come faster than a frequency of 3.3 Hz in this mode of operation. At each pulse a positive going signal will fire the Schmidt trigger which initiates A/D conversion. The program accepts the conversion and loops back through the conversion routine 24 additional times and finally com- putes the average. The Schmidt trigger is then reenabled and the program is readied for the next set of 25 points to be taken. Each conversion takes about 100 usec. Consequently, the total acquisition time is approximately 2.5 msec, thus leaving over 300 msec for the computer to average and process the data and get set for the next set of 25 points. Mode 2: Data acquisition at 1.0 cm“1 intervals In this mode, the triggering pulse comes at a frequency of 0.33 Hz. The method of data acquisition is identical 124 to that used in Mode 1. ADC Interface Since the largest analog signal available at the re- corder is approximately 10 mv, the signal must be amplified to give a significant ADC reading. This is accomplished through the use of a 141A operational amplifier with a gain of approximately 75. This supplies a 0.75 v signal to the ADC for a 10 mv signal at the recorder. CAUTION: The voltage input to the ADC must fall between -l.00 v and +1.00 v. Voltages outside these limits will result in meaningless data points. Since sufficient digital noise filtering cannot be ac- complished in the program, the amplifier is fitted with a low pass filter to eliminate the large amount of noise produced in the photon counting process. The triggering interface and its power supply are shown in Figure Al, while the ADC interface and its power supply are illustrated in Figure A2. OPERATION OF PROGRAM RAMAN Enabling 08/8 1. Mount the DECtape containing program RAMAN on the tape unit and roll about 10 ft. of tape onto the empty roll. 2. Place the tape unit in the WRITE LOCK and REMOTE mode and turn the teletype to LINE. 125 3A so E I V A 115v AC 5 6.3V - a) . ‘ 1 250““ MI 309 K 15" vomoe REGULATOR 5V 5V REGULATED POWER SUPPLY SPECTROMETER TRIGGERING INTERFACE Figure Al. Triggering interface and 5 volt regulated power supply. 126 Sufd 13300 100;“) ___—AAA MIA % Ho > + to ADC from RECORDER \ ’ I ADC INTERFACE -—1 A iIOOpfd ‘1 15V 01,1111 25,101 + 35v I I I +7 5 . T 35v ‘ IC 131010 _ 25111:) GNP l 35V fi -15\/ 117V AC uu'uuuua j—G < I k1 09999999090 000 _n '- \J < C iISV REGULATED 50 mA POWER SUPPLY Figure A2. Analog—digital converter and power supply. 127 Set the switch register on the CPU to 7470, 111 100 111 000. Press EXTD ADDR LOAD, ADDR LOAD, CLEAR, and CONTinue on the CPU in that order. If OS/8 is loaded, the teletype will respond with a dot (.). Turn on both the wavelength encoder and the amplifier. Plug the cable from the wavelength encoder into Schmidt trigger two (2). Check the voltage output range of the amplifier for a full-scale deflection of the recorder. This range must fall between —l.00v and +1.00v. The output voltage range of the amplifier may be changed by using the zero suppress control on the spectrometer. Plug the cable from the amplifier into channel zero (0) of the A/D converter. Running Program RAMAN On the teletype, type R RAMAN and RETURN. The teletype will respond with the question, "INITIAL WAVENUMBER?" Type in the initial wavenumber or tenth of a wave- number minus one (i.e., if the scan were to go 1 to 1250 cm—1 at 1.0 cm"1 intervals, from 950 cm— the user would type 949 or 949.0). The user then types the RETURN key. At this time the 128 spectrometer should be set at exactly 949.0 wave- numbers on the indicator dial, and the interval switch on the interface set to the proper position. 4. The teletype then quizzes, "FINAL WAVENUMBER?" 5. The user then types in the final wavenumber (i.e., 1250 or 1250.0) and then RETURN. 6. The teletype then replies, "INTERVAL BETWEEN DATA POINTS" 7. The user then types the interval between data points (i.e., 0.1 or 1.0) and then RETURN. CAUTION: If the interval between data points is 0.1 cm-l, the scan can only cover a maximum of 50 cm-l. Likewise, if the interval is 1.0 cm—l, the scan can cover a maximum of 500 cm-1. If these conditions are violated the program will give spurious results. 8. The user now begins the scan in a positive direction. 9. When the scan is finished the raw intensities are printed out and the teletype responds, "CHOOSE TWO WAVENUMBERS AS A BASELINE" 10. The user then types in two numbers to be used as a baseline (i.e., 950 or 950.0 RETURN, and then 1250 or 1250.0 RETURN). CAUTION: The user should never use the initial wavenumber (i.e., 949.0) as a point for baseline calculation. NOTE: The baseline calculated is a straight line and is automatically subtracted from the raw data. 11. 12. 13. 14. CAUTION: 129 The teletype responds, "DO YOU WISH TO INTEGRATE A PEAK?" The user replies with a zero (0) and RETURN for 22 and a one (1) and RETURN for yg_. If the answer is yes, the teletype will respond with, "CHOOSE YOUR INTEGRATION LIMITS" The user then types in the limits (i.e., 950 or 950.0 RETURN and 1250 or 1250.0 RETURN). The initial wavenumber (i.e., 949 or 949.0) may not be used as a limit. 15. 16. 17. 18. 19. Once the band has been integrated the program gives two more chances for integrating two additional spectral bands by repeating, "DO YOU WISH TO INTEGRATE A PEAK?" The user can reply as before. NOTE: If the user replies with a £2 to the question above, the question is not repeated. The teletype then responds, "THE AREA OF PEAK 1 IS number" "THE AREA OF PEAK 2 IS number" "THE AREA OF PEAK 3 IS number" NOTE: If only one band is integrated the data for peaks 2 and 3 will not appear. The teletype then responds, "TYPE A 1 FOR PEAKS OUTPUT:A ¢ FOR KINFIT OUTPUT" The user replies with a one (1) or a zero (0) and then RETURN. 20. 21. 22. 23. 24. 25. 26. 27. 130 If the KINFIT output was selected, the teletype asks, "WAVENUMBER VARIANCE?" The user types in the variance and RETURN. The teletype then quizzes, "INTENSITY VARIANCE?" The user again types in the variance and RETURN. The teletype responds with, "YOU HAVE 20 SECONDS TO TURN ON THE PUNCH!" At this time the user may punch a leader on the paper tape and prepare for the punching process. Upon completion of the punching process, the program will wait an additional 20 seconds to allow the user to punch a trailer on the paper tape. At the end of this waiting period the program will return to its beginning. NOTE: If an error is made during a run and the user wishes to restart the program, he toggles the bootstrap into the CPU as described in "Enabling OS/8". This returns computer control to MONITOR and the teletype will type a period (.). The user then types R RAMAN and proceeds again. TELETYPE 131 PROGRAM RAMAN FLONCHART START IINPUT SCAN PARAMETERS XSTAR XFINA XINTE MAKE ONE MK CONVERSION INPUT NAVENUMBER NW INTENSITY VARIANCES r___HHHIIHH_.H WAIT FOR PUNCH SETUP Figure A3. Flow chart AVERAGE AND POINTS STORE BEEN IN TAKEN DATA(J) LIST ROUGH DATA \ INPUT CALCULATE BASELINE BASELINE PARAMETERS AND SUBTRACT I INTEGRATE AND OUTPUT OUTPUT OUTPUT WAIT FOR » DATA ~ PUNCH SETUP for Program Raman. 803 (O mmmmmmmmmmt)mmmmmmmbimm (pm—J 132 P R0 G RAM RAI'IAN DIMENSION DATA<508>. 1975(100). AREAC3) MORE = l M = 2 I = 0 WHITE (1.1) FORMAT ('INITIAL MAVFMUMDERP') READ (1.2) XSTAR FORMAT(IF6-0) URITO (1.3) FORMAT (’FINAL RAUEMUMBERP') READ (1.4) XFINA FORMAT (IF6.O) WRITE (1.5) FORMAT ('INTERVAL arruasu DATA PJINTS?’) READ (1.6) XINTE FORMAT (IF6.O) INTER = (XFHM\-XSTAR)/KINTE DO 10 J = I. INTER M=I DATA(J) CPASE 22 START. CLA OLL /OLsHRs AOO.LINM a srarus Res 6135 TAD STOP 6130 CLA OLL TAD nor /SET CLOCK ENABLE REGISTER 6132 CLA OLL ASAIN. 6131 /HAS SCHMITT IRISOER FIRED? JMP AGAIN /N0. RAIT FOUR. CLA OLL 6532 /5TART couvsasron DONE. 6534 /OOMVERSION DONE? JMP DONE /N0. UAIT 6533 /vss. READ ADC INTO ACC DCA \I /ASSIGNMENT T3 VARIABLE SKP STOP. 7777 SKP ROT. 3302 IPTSCM) = I IF (M-25) 7.8.8 M = M + I JMP AOAIM Drs = 3 DO 630 10 = 1.25 015 = IPT5(IO) + 512 DATA = DATACJ) + PTS CONTINUE (3. 1'73 83 11 12 35 34 613 13 22 23 133 DATA = DATA(d)/2So CONTII’JUE WRITE(1.31) A = .‘(STA’I + .‘(INTES DO 83 L =1:INTER '.-1'RIT3(1.9) A. DATAH.) FORMAT(/. 11:113.]. I'ZX. 11:13.4) A = A + ‘(INTZ CONTINUE XJRITE< 1.11) FORMAT ('CIIOOSE T'w’O TJA‘JENUMBERS A3 A OASZLII‘JE’) READ (1.12) 53.315131 FORMAT (1176.13) READ (1.35) BASES FORMAT (1176.3) ICALl (BASEl-XSTAR)/XINTE ICAL2 = (BASE2-XSTAR)/XINTE BASH’ = DATA(ICAL1) BAS2Y == DATA(ICAL2) SLOPE = (BAS2Y-BASIY)/(BASE2-BASEl) YINTE = BASlY-SLOPE*BASEI A = XST R + XINTL‘ DO 131‘1 = 1. INTER Y = SLOPE’kA + YINTE.‘ DATACN) = DATAU‘J) - Y IF (DATA(N)) 34.63.63 DATAHJ) = Z A = A + .‘(INTE CONTINUE “171 TE (1. 23) FOQI’IATC’UO YOU WISH TO INTEGRATE A PZA1(?’) READ (1.214) KAI-15 FORMAT(1II) KANS = HANS + 1 GO TC) (26.19). HANS 'JRITE (1.23) FORMAT 'CHOOSE YOUR IIJTLSYHTIOZJ LIZ/TITS') '13AD(1.21) ALIEN FOWI~1AT<1F6~ 21) ’IEAI’)(1.76) AL1112 FORMAT( 1176.13) ILII-H (ALIH1-XSTAR)/XINTL 11.1312 (ALIEZ2-‘(STA'U/.\(IKJT:; SUM = .J [)0 9.5 I’ = 11.1511. ILI?-'1'.2 SUI-1 = 'SUI’. + Dr\T:’\(L‘I)*XI.‘-JT'L CONTINUES skiEAU-IJYL-L) = 511.”. IIO‘IE = >10 11?. +1 .30 TO 22 50 TO (342127327127). H xi) 5- (J m 29 28 342 343 344 32 163 104 185 126 345 69 67 68 31 346 167 33 32 138 169 347 349 999 Hr-o Hfi—l Mk» 134 MORE = NORE + 1 MARK = MORE - 2 DO 28 L = 1. MARK WRITE (1.29) L. AREACL) FORMATC'THE AREA OF PEAK ’.1I1.' IS '.1FIOo2) CONTINUE W31TE(1.343) FORMATC/l.’TYPE A 1 FOR PEAKS OUTPUT} A O FOR KINFIT OUTPUT') READ (1.344) NOUT ‘ FORMAT(111) IF (NOUT'Z) 33.33.345 WRITEC1.IZ3) FORMATC'WAVENUHBER VARIANCE7') READC1.104) HAV FORMATCIFéoO) 7JRITE(1.135) FORMATC'INTENSITY VARIANCE?’) READ(1.106) VAR FORMATC1F6oZ) WRITEC1.69) FORMAT(//.'YOU HAVE 23 SEC. TO TURN ON THE PUNCH!’.///) DO 68 MY = 1.733 M2 = O 90 67 NJ = 1. 230$ HZ = E-‘1Z + 1 CONTINUE CONTINUE IF (NOUT’Z) 346.346.347 FORMAT(//.' UAVENUMBER’.IOX.'INTENSITY') A = XSTAR + XINTE DO 32 M = 1.1NTER.2 AI = A + XINTE IF (M-INTER) 137.138.167 MM = M + l TRITE(1.33) A.MAV.DATA(M).VAR.A1.WAV.DATA(NM).VAR FORMATC4CFIZoO.FIZ.4)) A = A + 2. * XINTE CONTINUE GO TO 999 7.-13IT3(1.139) A. 'u-JA‘J. DATACDT). VAR FORMAT(2(F16.Z.1710.4)) GO TO 999 H?ITE(1.349) (DATA(M). M=1.INT£R) FO‘RMATC7(1F13~O)) DO 112 JY=1.7OO JZ=Z DO 113 IJ=1.2333 JZ=JZ+1 COIYTIIUIE COEFTIIVJC BO TO 33) £113 DATA(N) - IPTS(N) - AREA(N) - MORE - M XSTAR - XFINA - XINTE - XINTER - PTS - 1k in order 135 * PROGRAM RAMAN PARAMETERS real dimension used to store data points integer dimension used to store a set of 25 un- averaged points. real dimension used to store peak areas. integer variable used to index the peak areas. integer variable used to index the number of data points to be averaged. integer variable used to represent each data point which is transferred into IPTS(N). real input variable representing the initial wave- number. real input variable representing the final wave— number. real input variable representing the interval between data points. integer variable used to represent the number of data points to be taken. integer variable used as an index in data acquiSi- tion routine. real variable used to sum the 25 unaveraged data points and convert them to a real number. integer variable used as an index in the summing process. real variable used to index the appropriate wave- number upon output of data. of appearance in program listing BASE 1 BASE 2 ICALl and BASlY and SLOPE YINTE KANS ALIMl and ILIMl and SUM MARK NOUT WAV 136 integer variable used as an index when listing the rough data. real variable used to represent one point on the baseline. real variable used to represent one point on the baseline. ICALZ - integer variables used for indexing in the baseline calculation routine. BASZY - real variables representing the intensities at BASE l and BASE 2. real variable representing the slope of the base- line. real variable representing the Y-intercept of the baseline. real variable used to represent corrected intensity. integer variable used as an index for integrating peaks. ALIMZ - real variables used to represent the integration limits. ILIM2 - integer variables used to index the in- tensities at the integration limits. real variable representing a transient integral. integer variable used as an index for the output of the area of each peak. integer variable used as an index for KINFIT or PEAKS output. real variable representing the wavenumber variance. 137 VAR - real variable representing the intensity MY, MZ, and NJ - integer variables used as indices waiting routine Al — real variable used as wavenumber counter KINFIT output. MM — integer variable used to index intensity Jy, J2, and IJ - integer variables used as indices waiting routine. variance. in a in output. in second APPENDIX II DETERMINATION OF PHARMACOLOGICAL PROPERTIES OF 8-TERT-BUTYLPENTAMETHYLENETETRAZOLE AND 8-SEC-BUTYLPENTAMETHYLENETETRAZOLE AS QUOTED FROM COMMUNICATIONS RECEIVED FROM PROFESSOR WILLIAM E. STONE 138 DETERMINATION OF PHARMACOLOGICAL PROPERTIES OF 8-TERT-BUTYLPENTAMETHYLENETETRAZOLE AND 8-SEC-BUTYLPENTAMETHYLENETETRAZOLE AS QUOTED FROM COMMUNICATIONS RECEIVED FROM PROFESSOR WILLIAM E.STONE Professor William E. Stone, Department of Physiology, University of Wisconsin, Madison, Wisconsin, determined the pharmacological properties of the pentamethylenetetrazole derivatives. The tests were made on Swiss—Webster white mice, and the convulsants were administered intraperitoneally. Dose-response curves were calculated by the method of least squares, and for each compound the dose which induced a generalized seizure in 50% of the animals (CD50) was de- termined. The test results are given below. 8-tert-butylpentamethylenetetrazole The solubility of this compound in 0.85% sodium chloride solution appeared to be slightly less than 0.85 mg/ml. For these tests it was found advantageous to use 50% glycerol in water as the solvent. A solution which contained 1.6 mg of 8-Eert-butyl PMT per ml of solution was prepared. The compound dissolved completely on warming to 56°, but crystallization occurred overnight at room temperature. For lower doses, the solution that was used contained 1.2 mg/ml. There was also some crystallization from this solution 139 140 on standing overnight at room temperature. The CD50 was found to be 5.25 mg per kg of body weight, with 95% confidence limits of 3.97 to 6.93 (log CD50 = 0.719810.1207). The "approximate minimum convulsant dose" previously reported was 9 mg/kg (74). This falls at the point CD80 on the dose-response curve. In animals that had seizures, the latent periods were about 1 to 5 minutes. 8—sec—butylpentamethylenetetrazole The solubility of this compound seemed to be slightly greater than that of B—Egrt-butyl PMT in 0.85% sodium chloride solution, and considerably greater in 50% glycerol. The solution that was used contained 5 mg/ml of 8-sgg— butyl PMT in 50% glycerol. This amount dissolved completely on warming to 56°, but crystallized on cooling. The warm solution was used for injection. (It sometimes crystallized in the injection needle and plugged it.) The CD50 was found to be 48.6 mg/kg, with 95% confi- dence limits of 34.1 to 69.4 (log CD = 1.6870i0.1544). 50 The latent periods in animals having seizures were about 2 to 4 minutes. Pentamethylenetetrazole These results are included for comparison with the above tetrazoles. 141 The CD50 was 45.1 mg/kg, with 95% confidence limits of 40.3 to 50.2 (log CD 1.6542i0.0488). so= The "approximate minimum convulsant dose" previously reported was 50 mg/kg (74). This corresponds to the point CD on the dose—response curve. 69 In animals that had seizures, the latent periods were about 1.5 to 9 minutes. LITERATURE CITED 10. 11. 12. 13. 14. 15. 16. 17. 18. LITERATURE CITED J. F. Hinton and E. S. Amis, Chem. Rev., 11, 627 (1971). J. C. Evans and G. Y-S. Lo, J. Phys. Chem., _9, 3223 (1965) . W. F. Edgell, A. T. Watts, J. Lyford, and W. M. Risen, Jr., J. Amer. Chem. Soc., 88, 1815 (1966). W. F. Edgell, J. Lyford, R. Wright, W. M. Risen, Jr., and A. Watts, J. Amer. Chem. Soc., 92, 2240 (1970). B. W. Maxey and A. I. Popov, J. Amer. Chem. Soc., 82, 2230 (1967). B. W. Maxey and A. I. POpov, J. Amer. Chem. Soc., 21, 20 (1969). J. L. Wuepper and A. I. Popov, J. Amer. Chem. Soc., 91, 4352 (1969). J. L. Wuepper and A. I. Popov, J. Amer. Chem. Soc., 2_, 1493 (1970). M. K. Wong, W. J. McKinney, and A. I. Popov, J. Phys. Chem., 15, 56 (1971). M. K. Wong and A. I. Popov, J. Inorg. Nucl. Chem., 33, 1203 (1970). W. J. McKinney and A. I. Popov, J. Phys. Chem., 14, 535 (1970). P. R. Handy and A. I. Popov, Spectrochim. Acta., 28A, 1545 (1972). M. S. Greenberg, D. M. Wied, and A. I. Popov, Spectro— chim. Acta, 29A, 1927 (1973). R. H. Erlich, M. S. Greenberg, and A. I. Popov, Spectro- chim. Acta, 29A, 543 (1973). A. T. Tsatsas and W. M. Risen, Jr., J. Amer. Chem. Soc., 92, 1789 (1970). A. Regis and J. Corset, J. Chim. Phys. Physicochim. Biol., 62, 1508 (1972). C. Barker and J. Yarwood, J. C. S. Faraday II, 1322 (1975). A. T. Tsatsas and W. M. Risen, Jr., Chem. Phys. Lett., 1, 354 (1970). 142 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 32, 76 (1975). fig. 1466 (1974). 143 A. T. Tsatsas, R. W. stearns, and W. M. Risen, Jr., J. Amer. Chem. Soc., 24, 5247 (1972). Y. M. Cahen and A. I. Popov, J. Solution Chem., 4, 599 (1975). E. G. Hahn, J. A. Olander, and M. C. Day, J. Phys. Chem., 73, 3880 (1969). J. A. Olander and M. C. Day, J. Amer. Chem. Soc., 93, 3584 (1971). R. P. Taylor and I. D. Kuntz, Jr., J. Phys. Chem., 14, 4573 (1970). C. Lassigne and P. Baine, J. Phys. Chem., 15, 3188 W. F. Edgell, J. Lyford, A. Barbetta, C. I. Jose, J. Amer. Chem. Soc., 93, 6403 (1971). W. F. Edgell and J. Lyford, J. Amer. Chem. Soc., 93, 6407 (1971). W. F. Edgell and A. Barbetta, J. Amer. Chem. Soc., 26, 415 (1974). W. F. Edgell, W. R. Robinson, A. Barbetta, and D. P. Schussler, to be published. J. Borucka and Z. Kecki, Rocz. Chem., 45, 2133 (1971). J. Borucka and Z. Kecki, Rocz. Chem., 46, 1829 (1972). J. Borucka and Z. Kecki, J. Molec. Str., 26, 289 (1975). I. S. Perelygin and M. A. Klimchuk, Russ. J. Phys. Chem., .41, 1138 (1973). I. S. Perelygin and M. A. Klimchuk, Russ. J. Phys. Chem., 41, 1402 (1973). I. S. Perelygin and M. A. Klimchuk, Russ. J. Phys. Chem., 48, 363 (1974). I. S. Perelygin and M. A. Klimchuk, Russ. J. Phys. Chem., I. S. Perelygin and M. A. Klimchuk, Russ. J. Phys. Chem., M. Peleg, J. Phys. Chem., 26, 1019 (1972). 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 144 T. G. Chang and D. E. Irish, J. Phys. Chem., 11, 52 (1973). T. G. Chang and D. E. Irish, J. Solution Chem., 3, 161 (1974). T. G. Chang and D. E. Irish, J. Solution Chem., 3, 175 (1974). G. J. Janz and M. A. Muller, J. Solution Chem., 4, 285 (1975). D. E. Irish and R. V. Thorpe, Can. J. Chem., _3, 1414 (1975). K. R. Plowman and J. J. Lagowski, J. Phys. Chem., 18, 143 (1974). B. W. Maxey, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1968). J. L. Wuepper, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1969). M. K. Wong, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1971). P. R. Handy, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1972). A. Fratiello, R. E. Lee, and R. E. Schuster, Inorg. Chem., 9, 82 (1970). A. Fratiello, S. Peak, R. E. Schuster, and D. D. Davis, J. Phys. Chem., 14, 3730 (1971). A. Fratiello, D. D. Davis, S. Peak, and R. E. Schuster, Inorg. Chem., 10, 1627 (1971). E. Schaschel and M. C. Day, J. Amer. Chem. Soc., 20, 503 (1968). M. S. Greenberg, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1974). Y. M. Cahen, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1975). W. J. DeWitte, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1975). M. S. Greenberg and A. I. Popov, Spectrochim. Acta, 31A, 697 (1975). 14S 56. Y. M. Cahen, P. R. Handy, E. T. Roach, and A. I. Popov, J. Phys. Chem., 19, 80 (1975). 57. H. A. Berman and T. R. Stengle, J. Phys. Chem., 19, 1001 (1975). 58. W. J. DeWitte and A. I. Popov, J. Solution Chem., in press. 59. W. J. DeWitte and A. I. Popov, Inorg. Nucl. Chem. Lett., 12, 251 (1976). 60. Y. M. Cahen, J. L. Dye, and A. I. Popov, J. Phys. Chem., 19, 1289 (1975). 61. Y. M. Cahen, J. L. Dye, and A. I. Popov, J. Phys. Chem., 19, 1292 (1975). 62. E. Mei, J. L. Dye, and A. I. Popov, J. Amer. Chem. Soc., 98, 0000 (1976). 63. G. Foex, C. J. Gorter, and L. J. Smits, "Constantes Selectionnees, Diamagnetisme et Paramagnetisme, Relaxa— tion Paramagnetique", Masson & Cie Editeurs, Paris (1957). 64. J. L. Dye and V. A. Nicely, J. Chem. Educ., 48, 443 (1971). 65. M. S. Greenberg, R. L. Bodner, and A. I. Popov, J. Phys. Chem., 11, 2449 (1973). 66. U. Mayer and V. Gutmann, Struct. Bonding, 12, 113 (1972). '__ 67. V. Gutmann, "Coordination Chemistry in Nonaqueous Sol- vents", Springer-Verlag, Vienna (1968). 68. A. R. Davis, D. E. Irish, R. B. Roden, and A. J. Weer- heim, Appl. Spectrosc., 26, 384 (1972). 69. J. Bjerrum, "Metal Amine Formation in Aqueous Solutions", P. Haase and Son, Copenhagen (1941). 70. Y. M. Cahen, R. F. Beisel, and A. I. Popov, J. Inorg. Nucl. Chem., in press. 71. E. Olivera-Mandalla, Gazz. Chim. Ital., 44, 174 (1914). 72. D. W. Moore and A. G. Whittaker, J. Amer. Chem. Soc., 2, 5007 (1960). ___-I— 73. J. B. Lounsbury, J. Phys. Chem., 67, 721 (1963). 146 74. . E. Stone, Pharmacology, 3, 367 (1970). W 75. A. Dister, J. Pharm. Belg., 3, 190 (1948). A 76. . I. Popov, Coord. Chem. Rev., 4, 463 (1969). 77. L. K. Sharp. J. Pharm. Pharmacol., 4, 52 (1952). 78. T. E. V. Horsley, Analyst, 14, 308 (1946). 79. E. Anderson, M. Fors, and J. Lindgren, Acta Chem. Scand., 44, 1957 (1960). 80. A. I. Popov and J. C. Marshall, J. Inorg. Nucl. Chem., 12' 340 (1961). 81. A. I. Popov and J. C. Marshall, J. Inorg. Nucl. Chem., .24, 1667 (1962). 82. T. Beyrich and G. Schlaak, Pharmazie, 34, 152 (1969). 83. R. A. Daoust, J. Pharm. Sci., 52, 642 (1963). 84. J. W. Turczan and B. A. Goldwitz, J. Pharm. Sci., 44, 1309 (1972). 85. E. Haywood, M. W. Horner, and H. J. Rylance, Analyst, P. 23, 711 (1967). 86. K. Guven, Eczacilik Bu1., 13' 93 (1970). 87. H. Kawamoto, Kumamoto Medical J., 44, 69 (1962). 88. F. Marcucci, M. L. Airoldi, E. Mussini, and S. Garattini, Eur. J. Pharmacol., 44, 219 (1971). 89. J. T. Stewart and J. L. Story, J. Pharm. Sci., 44, 1651 (1972). 90. C. Cardini, G. Cavazzutti, and V. Quercia, Boll. Chim. Farm., 109, 333 (1970). 91. A. I. Popov and R. D. Holm, J. Phys. Chem., 44, 158 (1962). 92. F. W. Schueler, S. C. Wang, R. M. Featherstone, and E. G. Gross, J. Pharmac. Exp. Ther., 9;, 266 (1949). 93. R. H. Erlich and A. I. Popov, J. Phys. Chem., 14, 338 (1970). 94. A. S. Buchanan, G. M. Laycock, and A. Shulman, Eur. J. Pharmacol., Z, 60 (1969). 147 95. F. Modrzejewski and J. Rygiel, Acta Pol. Pharm., 32, 175 (1972). 96. G. J. Gross and D. M. Woodbury, J. Pharmac. Exp. Ther., 181, 257 (1972). 97. H. Meyer, Arch. Exptl. Pathol. Pharmakol., 42, 110 (1899). _— 98. E. Overton, "Studien uber die Narkose", Fischer, Jena, Germany (1901). 99. C. Hansch, Accounts Chem. Res., 3' 232 (1969). 100. F. Helmer, K. Kichs, and C. Hansch, Biochemistry, Z, 2858 (1968). 101. A. Leo, C. Hansch, and D. Elkins, Chem. Rev., 71, 525 (1971). ——' 102. F. M. D'Itri, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1968). 103. F. M. D'Itri and A. I. Popov, J. Amer. Chem. Soc., 90, 6476 (1968). _— 104. W. T. Somerville and E. T. Theimer, U.S. Patent 2,840,599 (June 24, 1958), C.A. 53, 18265i. 105. J.\nn1Braun, Ann., 472, 70 (1929). 106. L. T. Sandboom, Org. Synthesis, Coll. Vol. I, p. 340 (1941). 107. E. K. Harvill, C. W. Roberts, and R. M. Herbst, J. Org. Chem., 45, 58 (1950). 108. D. P. Shoemaker and C. W. Garland, "Experiments in Physical Chemistry", p. 132, McGraw—Hill, New York (1967). 109. J. C. Giddings, J. Chem. Educ., 39, 569 (1962). 110. H. M. McNair and E. J. Bonelli, "Basic Gas Chromatography", Varian Aerograph, Walnut Creek, California, p. 127 (1969). 111. R. H. Erlich, Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1971). 112. E. G. Gross and R. M. Featherstone, J. Pharmac. Exp. Ther., 81, 291 (1946).